id
large_stringlengths 9
16
| submitter
large_stringlengths 3
64
⌀ | authors
large_stringlengths 5
20.2k
| title
large_stringlengths 5
245
| comments
large_stringlengths 1
482
⌀ | journal-ref
large_stringlengths 4
382
⌀ | doi
large_stringlengths 9
151
⌀ | report-no
large_stringlengths 2
509
⌀ | categories
large_stringlengths 5
78
| license
large_stringclasses 9
values | abstract
large_stringlengths 83
3.55k
| versions
large listlengths 1
21
| update_date
timestamp[ms]date 2007-05-23 00:00:00
2025-05-30 00:00:00
| authors_parsed
large listlengths 1
1.35k
| prompt
large_stringlengths 166
3.64k
| label
stringclasses 2
values | prob
float64 0.5
0.73
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2305.03052 | Basile Van Hoorick | Basile Van Hoorick, Pavel Tokmakov, Simon Stent, Jie Li, Carl Vondrick | Tracking through Containers and Occluders in the Wild | Accepted at CVPR 2023. Project webpage is available at:
https://tcow.cs.columbia.edu/ | null | null | null | cs.CV cs.AI cs.LG cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Tracking objects with persistence in cluttered and dynamic environments
remains a difficult challenge for computer vision systems. In this paper, we
introduce $\textbf{TCOW}$, a new benchmark and model for visual tracking
through heavy occlusion and containment. We set up a task where the goal is to,
given a video sequence, segment both the projected extent of the target object,
as well as the surrounding container or occluder whenever one exists. To study
this task, we create a mixture of synthetic and annotated real datasets to
support both supervised learning and structured evaluation of model performance
under various forms of task variation, such as moving or nested containment. We
evaluate two recent transformer-based video models and find that while they can
be surprisingly capable of tracking targets under certain settings of task
variation, there remains a considerable performance gap before we can claim a
tracking model to have acquired a true notion of object permanence.
| [
{
"version": "v1",
"created": "Thu, 4 May 2023 17:59:58 GMT"
}
] | 2023-05-05T00:00:00 | [
[
"Van Hoorick",
"Basile",
""
],
[
"Tokmakov",
"Pavel",
""
],
[
"Stent",
"Simon",
""
],
[
"Li",
"Jie",
""
],
[
"Vondrick",
"Carl",
""
]
] | TITLE: Tracking through Containers and Occluders in the Wild
ABSTRACT: Tracking objects with persistence in cluttered and dynamic environments
remains a difficult challenge for computer vision systems. In this paper, we
introduce $\textbf{TCOW}$, a new benchmark and model for visual tracking
through heavy occlusion and containment. We set up a task where the goal is to,
given a video sequence, segment both the projected extent of the target object,
as well as the surrounding container or occluder whenever one exists. To study
this task, we create a mixture of synthetic and annotated real datasets to
support both supervised learning and structured evaluation of model performance
under various forms of task variation, such as moving or nested containment. We
evaluate two recent transformer-based video models and find that while they can
be surprisingly capable of tracking targets under certain settings of task
variation, there remains a considerable performance gap before we can claim a
tracking model to have acquired a true notion of object permanence.
| new_dataset | 0.715405 |
2007.03746 | Dongrui Wu | Dongrui Wu and Xue Jiang and Ruimin Peng and Wanzeng Kong and Jian
Huang and Zhigang Zeng | Transfer Learning for Motor Imagery Based Brain-Computer Interfaces: A
Complete Pipeline | null | Neural Networks, 153:235-253, 2022 | 10.1016/j.neunet.2022.06.008 | null | eess.SP cs.HC cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Transfer learning (TL) has been widely used in motor imagery (MI) based
brain-computer interfaces (BCIs) to reduce the calibration effort for a new
subject, and demonstrated promising performance. While a closed-loop MI-based
BCI system, after electroencephalogram (EEG) signal acquisition and temporal
filtering, includes spatial filtering, feature engineering, and classification
blocks before sending out the control signal to an external device, previous
approaches only considered TL in one or two such components. This paper
proposes that TL could be considered in all three components (spatial
filtering, feature engineering, and classification) of MI-based BCIs.
Furthermore, it is also very important to specifically add a data alignment
component before spatial filtering to make the data from different subjects
more consistent, and hence to facilitate subsequential TL. Offline calibration
experiments on two MI datasets verified our proposal. Especially, integrating
data alignment and sophisticated TL approaches can significantly improve the
classification performance, and hence greatly reduces the calibration effort.
| [
{
"version": "v1",
"created": "Fri, 3 Jul 2020 23:44:21 GMT"
},
{
"version": "v2",
"created": "Thu, 8 Oct 2020 20:56:41 GMT"
},
{
"version": "v3",
"created": "Fri, 22 Jan 2021 20:37:14 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Wu",
"Dongrui",
""
],
[
"Jiang",
"Xue",
""
],
[
"Peng",
"Ruimin",
""
],
[
"Kong",
"Wanzeng",
""
],
[
"Huang",
"Jian",
""
],
[
"Zeng",
"Zhigang",
""
]
] | TITLE: Transfer Learning for Motor Imagery Based Brain-Computer Interfaces: A
Complete Pipeline
ABSTRACT: Transfer learning (TL) has been widely used in motor imagery (MI) based
brain-computer interfaces (BCIs) to reduce the calibration effort for a new
subject, and demonstrated promising performance. While a closed-loop MI-based
BCI system, after electroencephalogram (EEG) signal acquisition and temporal
filtering, includes spatial filtering, feature engineering, and classification
blocks before sending out the control signal to an external device, previous
approaches only considered TL in one or two such components. This paper
proposes that TL could be considered in all three components (spatial
filtering, feature engineering, and classification) of MI-based BCIs.
Furthermore, it is also very important to specifically add a data alignment
component before spatial filtering to make the data from different subjects
more consistent, and hence to facilitate subsequential TL. Offline calibration
experiments on two MI datasets verified our proposal. Especially, integrating
data alignment and sophisticated TL approaches can significantly improve the
classification performance, and hence greatly reduces the calibration effort.
| no_new_dataset | 0.709466 |
2102.03310 | Michal Ciszewski | Micha{\l} Ciszewski, Jakob S\"ohl, Geurt Jongbloed | Improving state estimation through projection post-processing for
activity recognition with application to football | This preprint has not undergone peer review (when applicable) or any
post-submission improvements or corrections. The Version of Record of this
article is published in Statistical Methods & Applications, and is available
online at https://doi.org/10.1007/s10260-023-00696-z | Stat Methods Appl (2023) | 10.1007/s10260-023-00696-z | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | The past decade has seen an increased interest in human activity recognition
based on sensor data. Most often, the sensor data come unannotated, creating
the need for fast labelling methods. For assessing the quality of the
labelling, an appropriate performance measure has to be chosen. Our main
contribution is a novel post-processing method for activity recognition. It
improves the accuracy of the classification methods by correcting for
unrealistic short activities in the estimate. We also propose a new performance
measure, the Locally Time-Shifted Measure (LTS measure), which addresses
uncertainty in the times of state changes. The effectiveness of the
post-processing method is evaluated, using the novel LTS measure, on the basis
of a simulated dataset and a real application on sensor data from football. The
simulation study is also used to discuss the choice of the parameters of the
post-processing method and the LTS measure.
| [
{
"version": "v1",
"created": "Fri, 5 Feb 2021 17:32:39 GMT"
},
{
"version": "v2",
"created": "Thu, 10 Jun 2021 09:43:01 GMT"
},
{
"version": "v3",
"created": "Fri, 2 Sep 2022 10:27:28 GMT"
},
{
"version": "v4",
"created": "Tue, 2 May 2023 19:56:30 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Ciszewski",
"Michał",
""
],
[
"Söhl",
"Jakob",
""
],
[
"Jongbloed",
"Geurt",
""
]
] | TITLE: Improving state estimation through projection post-processing for
activity recognition with application to football
ABSTRACT: The past decade has seen an increased interest in human activity recognition
based on sensor data. Most often, the sensor data come unannotated, creating
the need for fast labelling methods. For assessing the quality of the
labelling, an appropriate performance measure has to be chosen. Our main
contribution is a novel post-processing method for activity recognition. It
improves the accuracy of the classification methods by correcting for
unrealistic short activities in the estimate. We also propose a new performance
measure, the Locally Time-Shifted Measure (LTS measure), which addresses
uncertainty in the times of state changes. The effectiveness of the
post-processing method is evaluated, using the novel LTS measure, on the basis
of a simulated dataset and a real application on sensor data from football. The
simulation study is also used to discuss the choice of the parameters of the
post-processing method and the LTS measure.
| no_new_dataset | 0.711249 |
2107.05729 | Yicheng Fei | Yicheng Fei, Xaq Pitkow | Generalization of graph network inferences in higher-order graphical
models | 14 pages, 5 figures | null | null | null | cs.AI cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | Probabilistic graphical models provide a powerful tool to describe complex
statistical structure, with many real-world applications in science and
engineering from controlling robotic arms to understanding neuronal
computations. A major challenge for these graphical models is that inferences
such as marginalization are intractable for general graphs. These inferences
are often approximated by a distributed message-passing algorithm such as
Belief Propagation, which does not always perform well on graphs with cycles,
nor can it always be easily specified for complex continuous probability
distributions. Such difficulties arise frequently in expressive graphical
models that include intractable higher-order interactions. In this paper we
define the Recurrent Factor Graph Neural Network (RF-GNN) to achieve fast
approximate inference on graphical models that involve many-variable
interactions. Experimental results on several families of graphical models
demonstrate the out-of-distribution generalization capability of our method to
different sized graphs, and indicate the domain in which our method outperforms
Belief Propagation (BP). Moreover, we test the RF-GNN on a real-world
Low-Density Parity-Check dataset as a benchmark along with other baseline
models including BP variants and other GNN methods. Overall we find that
RF-GNNs outperform other methods under high noise levels.
| [
{
"version": "v1",
"created": "Mon, 12 Jul 2021 20:51:27 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 03:25:53 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Fei",
"Yicheng",
""
],
[
"Pitkow",
"Xaq",
""
]
] | TITLE: Generalization of graph network inferences in higher-order graphical
models
ABSTRACT: Probabilistic graphical models provide a powerful tool to describe complex
statistical structure, with many real-world applications in science and
engineering from controlling robotic arms to understanding neuronal
computations. A major challenge for these graphical models is that inferences
such as marginalization are intractable for general graphs. These inferences
are often approximated by a distributed message-passing algorithm such as
Belief Propagation, which does not always perform well on graphs with cycles,
nor can it always be easily specified for complex continuous probability
distributions. Such difficulties arise frequently in expressive graphical
models that include intractable higher-order interactions. In this paper we
define the Recurrent Factor Graph Neural Network (RF-GNN) to achieve fast
approximate inference on graphical models that involve many-variable
interactions. Experimental results on several families of graphical models
demonstrate the out-of-distribution generalization capability of our method to
different sized graphs, and indicate the domain in which our method outperforms
Belief Propagation (BP). Moreover, we test the RF-GNN on a real-world
Low-Density Parity-Check dataset as a benchmark along with other baseline
models including BP variants and other GNN methods. Overall we find that
RF-GNNs outperform other methods under high noise levels.
| no_new_dataset | 0.708213 |
2112.05807 | Hannes Westermann | Hannes Westermann, Jaromir Savelka, Vern R. Walker, Kevin D. Ashley,
Karim Benyekhlef | Computer-Assisted Creation of Boolean Search Rules for Text
Classification in the Legal Domain | null | Frontiers in Artificial Intelligence and Applications, Volume 322:
Legal Knowledge and Information Systems, 2019, pp. 123 - 132 | 10.3233/FAIA190313 | null | cs.LG cs.AI cs.CL cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we present a method of building strong, explainable
classifiers in the form of Boolean search rules. We developed an interactive
environment called CASE (Computer Assisted Semantic Exploration) which exploits
word co-occurrence to guide human annotators in selection of relevant search
terms. The system seamlessly facilitates iterative evaluation and improvement
of the classification rules. The process enables the human annotators to
leverage the benefits of statistical information while incorporating their
expert intuition into the creation of such rules. We evaluate classifiers
created with our CASE system on 4 datasets, and compare the results to machine
learning methods, including SKOPE rules, Random forest, Support Vector Machine,
and fastText classifiers. The results drive the discussion on trade-offs
between superior compactness, simplicity, and intuitiveness of the Boolean
search rules versus the better performance of state-of-the-art machine learning
models for text classification.
| [
{
"version": "v1",
"created": "Fri, 10 Dec 2021 19:53:41 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Westermann",
"Hannes",
""
],
[
"Savelka",
"Jaromir",
""
],
[
"Walker",
"Vern R.",
""
],
[
"Ashley",
"Kevin D.",
""
],
[
"Benyekhlef",
"Karim",
""
]
] | TITLE: Computer-Assisted Creation of Boolean Search Rules for Text
Classification in the Legal Domain
ABSTRACT: In this paper, we present a method of building strong, explainable
classifiers in the form of Boolean search rules. We developed an interactive
environment called CASE (Computer Assisted Semantic Exploration) which exploits
word co-occurrence to guide human annotators in selection of relevant search
terms. The system seamlessly facilitates iterative evaluation and improvement
of the classification rules. The process enables the human annotators to
leverage the benefits of statistical information while incorporating their
expert intuition into the creation of such rules. We evaluate classifiers
created with our CASE system on 4 datasets, and compare the results to machine
learning methods, including SKOPE rules, Random forest, Support Vector Machine,
and fastText classifiers. The results drive the discussion on trade-offs
between superior compactness, simplicity, and intuitiveness of the Boolean
search rules versus the better performance of state-of-the-art machine learning
models for text classification.
| no_new_dataset | 0.712389 |
2202.10847 | Francisca Vasconcelos | Francisca Vasconcelos, Bobby He, Nalini Singh, Yee Whye Teh | UncertaINR: Uncertainty Quantification of End-to-End Implicit Neural
Representations for Computed Tomography | Published in the Transactions on Machine Learning Research (TMLR)
April 2023 [https://openreview.net/forum?id=jdGMBgYvfX] | null | null | null | eess.IV cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Implicit neural representations (INRs) have achieved impressive results for
scene reconstruction and computer graphics, where their performance has
primarily been assessed on reconstruction accuracy. As INRs make their way into
other domains, where model predictions inform high-stakes decision-making,
uncertainty quantification of INR inference is becoming critical. To that end,
we study a Bayesian reformulation of INRs, UncertaINR, in the context of
computed tomography, and evaluate several Bayesian deep learning
implementations in terms of accuracy and calibration. We find that they achieve
well-calibrated uncertainty, while retaining accuracy competitive with other
classical, INR-based, and CNN-based reconstruction techniques. Contrary to
common intuition in the Bayesian deep learning literature, we find that INRs
obtain the best calibration with computationally efficient Monte Carlo dropout,
outperforming Hamiltonian Monte Carlo and deep ensembles. Moreover, in contrast
to the best-performing prior approaches, UncertaINR does not require a large
training dataset, but only a handful of validation images.
| [
{
"version": "v1",
"created": "Tue, 22 Feb 2022 12:19:03 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Jun 2022 11:25:29 GMT"
},
{
"version": "v3",
"created": "Tue, 2 May 2023 20:59:18 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Vasconcelos",
"Francisca",
""
],
[
"He",
"Bobby",
""
],
[
"Singh",
"Nalini",
""
],
[
"Teh",
"Yee Whye",
""
]
] | TITLE: UncertaINR: Uncertainty Quantification of End-to-End Implicit Neural
Representations for Computed Tomography
ABSTRACT: Implicit neural representations (INRs) have achieved impressive results for
scene reconstruction and computer graphics, where their performance has
primarily been assessed on reconstruction accuracy. As INRs make their way into
other domains, where model predictions inform high-stakes decision-making,
uncertainty quantification of INR inference is becoming critical. To that end,
we study a Bayesian reformulation of INRs, UncertaINR, in the context of
computed tomography, and evaluate several Bayesian deep learning
implementations in terms of accuracy and calibration. We find that they achieve
well-calibrated uncertainty, while retaining accuracy competitive with other
classical, INR-based, and CNN-based reconstruction techniques. Contrary to
common intuition in the Bayesian deep learning literature, we find that INRs
obtain the best calibration with computationally efficient Monte Carlo dropout,
outperforming Hamiltonian Monte Carlo and deep ensembles. Moreover, in contrast
to the best-performing prior approaches, UncertaINR does not require a large
training dataset, but only a handful of validation images.
| no_new_dataset | 0.709573 |
2204.12399 | Ankit Pensia | Ilias Diakonikolas, Daniel M. Kane, Ankit Pensia, Thanasis Pittas | Streaming Algorithms for High-Dimensional Robust Statistics | null | null | null | null | cs.DS cs.LG math.ST stat.ML stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study high-dimensional robust statistics tasks in the streaming model. A
recent line of work obtained computationally efficient algorithms for a range
of high-dimensional robust estimation tasks. Unfortunately, all previous
algorithms require storing the entire dataset, incurring memory at least
quadratic in the dimension. In this work, we develop the first efficient
streaming algorithms for high-dimensional robust statistics with near-optimal
memory requirements (up to logarithmic factors). Our main result is for the
task of high-dimensional robust mean estimation in (a strengthening of) Huber's
contamination model. We give an efficient single-pass streaming algorithm for
this task with near-optimal error guarantees and space complexity nearly-linear
in the dimension. As a corollary, we obtain streaming algorithms with
near-optimal space complexity for several more complex tasks, including robust
covariance estimation, robust regression, and more generally robust stochastic
optimization.
| [
{
"version": "v1",
"created": "Tue, 26 Apr 2022 15:57:07 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 17:59:26 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Diakonikolas",
"Ilias",
""
],
[
"Kane",
"Daniel M.",
""
],
[
"Pensia",
"Ankit",
""
],
[
"Pittas",
"Thanasis",
""
]
] | TITLE: Streaming Algorithms for High-Dimensional Robust Statistics
ABSTRACT: We study high-dimensional robust statistics tasks in the streaming model. A
recent line of work obtained computationally efficient algorithms for a range
of high-dimensional robust estimation tasks. Unfortunately, all previous
algorithms require storing the entire dataset, incurring memory at least
quadratic in the dimension. In this work, we develop the first efficient
streaming algorithms for high-dimensional robust statistics with near-optimal
memory requirements (up to logarithmic factors). Our main result is for the
task of high-dimensional robust mean estimation in (a strengthening of) Huber's
contamination model. We give an efficient single-pass streaming algorithm for
this task with near-optimal error guarantees and space complexity nearly-linear
in the dimension. As a corollary, we obtain streaming algorithms with
near-optimal space complexity for several more complex tasks, including robust
covariance estimation, robust regression, and more generally robust stochastic
optimization.
| no_new_dataset | 0.710245 |
2205.00470 | Konstantin D Pandl | Konstantin D. Pandl, Florian Leiser, Scott Thiebes, Ali Sunyaev | Reward Systems for Trustworthy Medical Federated Learning | 10 pages | null | null | null | cs.LG cs.CY cs.GT | http://creativecommons.org/licenses/by-sa/4.0/ | Federated learning (FL) has received high interest from researchers and
practitioners to train machine learning (ML) models for healthcare. Ensuring
the trustworthiness of these models is essential. Especially bias, defined as a
disparity in the model's predictive performance across different subgroups, may
cause unfairness against specific subgroups, which is an undesired phenomenon
for trustworthy ML models. In this research, we address the question to which
extent bias occurs in medical FL and how to prevent excessive bias through
reward systems. We first evaluate how to measure the contributions of
institutions toward predictive performance and bias in cross-silo medical FL
with a Shapley value approximation method. In a second step, we design
different reward systems incentivizing contributions toward high predictive
performance or low bias. We then propose a combined reward system that
incentivizes contributions toward both. We evaluate our work using multiple
medical chest X-ray datasets focusing on patient subgroups defined by patient
sex and age. Our results show that we can successfully measure contributions
toward bias, and an integrated reward system successfully incentivizes
contributions toward a well-performing model with low bias. While the
partitioning of scans only slightly influences the overall bias, institutions
with data predominantly from one subgroup introduce a favorable bias for this
subgroup. Our results indicate that reward systems, which focus on predictive
performance only, can transfer model bias against patients to an institutional
level. Our work helps researchers and practitioners design reward systems for
FL with well-aligned incentives for trustworthy ML.
| [
{
"version": "v1",
"created": "Sun, 1 May 2022 14:01:33 GMT"
},
{
"version": "v2",
"created": "Tue, 2 May 2023 19:27:42 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Pandl",
"Konstantin D.",
""
],
[
"Leiser",
"Florian",
""
],
[
"Thiebes",
"Scott",
""
],
[
"Sunyaev",
"Ali",
""
]
] | TITLE: Reward Systems for Trustworthy Medical Federated Learning
ABSTRACT: Federated learning (FL) has received high interest from researchers and
practitioners to train machine learning (ML) models for healthcare. Ensuring
the trustworthiness of these models is essential. Especially bias, defined as a
disparity in the model's predictive performance across different subgroups, may
cause unfairness against specific subgroups, which is an undesired phenomenon
for trustworthy ML models. In this research, we address the question to which
extent bias occurs in medical FL and how to prevent excessive bias through
reward systems. We first evaluate how to measure the contributions of
institutions toward predictive performance and bias in cross-silo medical FL
with a Shapley value approximation method. In a second step, we design
different reward systems incentivizing contributions toward high predictive
performance or low bias. We then propose a combined reward system that
incentivizes contributions toward both. We evaluate our work using multiple
medical chest X-ray datasets focusing on patient subgroups defined by patient
sex and age. Our results show that we can successfully measure contributions
toward bias, and an integrated reward system successfully incentivizes
contributions toward a well-performing model with low bias. While the
partitioning of scans only slightly influences the overall bias, institutions
with data predominantly from one subgroup introduce a favorable bias for this
subgroup. Our results indicate that reward systems, which focus on predictive
performance only, can transfer model bias against patients to an institutional
level. Our work helps researchers and practitioners design reward systems for
FL with well-aligned incentives for trustworthy ML.
| no_new_dataset | 0.709007 |
2205.05656 | Hang Dong | Hang Dong, V\'ictor Su\'arez-Paniagua, Huayu Zhang, Minhong Wang,
Arlene Casey, Emma Davidson, Jiaoyan Chen, Beatrice Alex, William Whiteley,
Honghan Wu | Ontology-Driven and Weakly Supervised Rare Disease Identification from
Clinical Notes | Accepted for BMC Medical Informatics and Decision Making, structured
abstract in full text, 16 pages, 4 figures (and extra 7 pages, 1 figure in
the supplementary material) | null | 10.1186/s12911-023-02181-9 | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Computational text phenotyping is the practice of identifying patients with
certain disorders and traits from clinical notes. Rare diseases are challenging
to be identified due to few cases available for machine learning and the need
for data annotation from domain experts. We propose a method using ontologies
and weak supervision, with recent pre-trained contextual representations from
Bi-directional Transformers (e.g. BERT). The ontology-based framework includes
two steps: (i) Text-to-UMLS, extracting phenotypes by contextually linking
mentions to concepts in Unified Medical Language System (UMLS), with a Named
Entity Recognition and Linking (NER+L) tool, SemEHR, and weak supervision with
customised rules and contextual mention representation; (ii) UMLS-to-ORDO,
matching UMLS concepts to rare diseases in Orphanet Rare Disease Ontology
(ORDO). The weakly supervised approach is proposed to learn a phenotype
confirmation model to improve Text-to-UMLS linking, without annotated data from
domain experts. We evaluated the approach on three clinical datasets, MIMIC-III
discharge summaries, MIMIC-III radiology reports, and NHS Tayside brain imaging
reports from two institutions in the US and the UK, with annotations. The
improvements in the precision were pronounced (by over 30% to 50% absolute
score for Text-to-UMLS linking), with almost no loss of recall compared to the
existing NER+L tool, SemEHR. Results on radiology reports from MIMIC-III and
NHS Tayside were consistent with the discharge summaries. The overall pipeline
processing clinical notes can extract rare disease cases, mostly uncaptured in
structured data (manually assigned ICD codes). We discuss the usefulness of the
weak supervision approach and propose directions for future studies.
| [
{
"version": "v1",
"created": "Wed, 11 May 2022 17:38:24 GMT"
},
{
"version": "v2",
"created": "Wed, 3 Aug 2022 18:42:55 GMT"
},
{
"version": "v3",
"created": "Fri, 9 Sep 2022 16:33:53 GMT"
},
{
"version": "v4",
"created": "Tue, 28 Feb 2023 20:58:22 GMT"
},
{
"version": "v5",
"created": "Wed, 3 May 2023 06:55:48 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Dong",
"Hang",
""
],
[
"Suárez-Paniagua",
"Víctor",
""
],
[
"Zhang",
"Huayu",
""
],
[
"Wang",
"Minhong",
""
],
[
"Casey",
"Arlene",
""
],
[
"Davidson",
"Emma",
""
],
[
"Chen",
"Jiaoyan",
""
],
[
"Alex",
"Beatrice",
""
],
[
"Whiteley",
"William",
""
],
[
"Wu",
"Honghan",
""
]
] | TITLE: Ontology-Driven and Weakly Supervised Rare Disease Identification from
Clinical Notes
ABSTRACT: Computational text phenotyping is the practice of identifying patients with
certain disorders and traits from clinical notes. Rare diseases are challenging
to be identified due to few cases available for machine learning and the need
for data annotation from domain experts. We propose a method using ontologies
and weak supervision, with recent pre-trained contextual representations from
Bi-directional Transformers (e.g. BERT). The ontology-based framework includes
two steps: (i) Text-to-UMLS, extracting phenotypes by contextually linking
mentions to concepts in Unified Medical Language System (UMLS), with a Named
Entity Recognition and Linking (NER+L) tool, SemEHR, and weak supervision with
customised rules and contextual mention representation; (ii) UMLS-to-ORDO,
matching UMLS concepts to rare diseases in Orphanet Rare Disease Ontology
(ORDO). The weakly supervised approach is proposed to learn a phenotype
confirmation model to improve Text-to-UMLS linking, without annotated data from
domain experts. We evaluated the approach on three clinical datasets, MIMIC-III
discharge summaries, MIMIC-III radiology reports, and NHS Tayside brain imaging
reports from two institutions in the US and the UK, with annotations. The
improvements in the precision were pronounced (by over 30% to 50% absolute
score for Text-to-UMLS linking), with almost no loss of recall compared to the
existing NER+L tool, SemEHR. Results on radiology reports from MIMIC-III and
NHS Tayside were consistent with the discharge summaries. The overall pipeline
processing clinical notes can extract rare disease cases, mostly uncaptured in
structured data (manually assigned ICD codes). We discuss the usefulness of the
weak supervision approach and propose directions for future studies.
| no_new_dataset | 0.713849 |
2205.11332 | Liang Zeng | Liang Zeng, Lanqing Li, Ziqi Gao, Peilin Zhao, Jian Li | ImGCL: Revisiting Graph Contrastive Learning on Imbalanced Node
Classification | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph contrastive learning (GCL) has attracted a surge of attention due to
its superior performance for learning node/graph representations without
labels. However, in practice, the underlying class distribution of unlabeled
nodes for the given graph is usually imbalanced. This highly imbalanced class
distribution inevitably deteriorates the quality of learned node
representations in GCL. Indeed, we empirically find that most state-of-the-art
GCL methods cannot obtain discriminative representations and exhibit poor
performance on imbalanced node classification. Motivated by this observation,
we propose a principled GCL framework on Imbalanced node classification
(ImGCL), which automatically and adaptively balances the representations
learned from GCL without labels. Specifically, we first introduce the online
clustering based progressively balanced sampling (PBS) method with theoretical
rationale, which balances the training sets based on pseudo-labels obtained
from learned representations in GCL. We then develop the node centrality based
PBS method to better preserve the intrinsic structure of graphs, by upweighting
the important nodes of the given graph. Extensive experiments on multiple
imbalanced graph datasets and imbalanced settings demonstrate the effectiveness
of our proposed framework, which significantly improves the performance of the
recent state-of-the-art GCL methods. Further experimental ablations and
analyses show that the ImGCL framework consistently improves the representation
quality of nodes in under-represented (tail) classes.
| [
{
"version": "v1",
"created": "Mon, 23 May 2022 14:23:36 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 06:00:40 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Zeng",
"Liang",
""
],
[
"Li",
"Lanqing",
""
],
[
"Gao",
"Ziqi",
""
],
[
"Zhao",
"Peilin",
""
],
[
"Li",
"Jian",
""
]
] | TITLE: ImGCL: Revisiting Graph Contrastive Learning on Imbalanced Node
Classification
ABSTRACT: Graph contrastive learning (GCL) has attracted a surge of attention due to
its superior performance for learning node/graph representations without
labels. However, in practice, the underlying class distribution of unlabeled
nodes for the given graph is usually imbalanced. This highly imbalanced class
distribution inevitably deteriorates the quality of learned node
representations in GCL. Indeed, we empirically find that most state-of-the-art
GCL methods cannot obtain discriminative representations and exhibit poor
performance on imbalanced node classification. Motivated by this observation,
we propose a principled GCL framework on Imbalanced node classification
(ImGCL), which automatically and adaptively balances the representations
learned from GCL without labels. Specifically, we first introduce the online
clustering based progressively balanced sampling (PBS) method with theoretical
rationale, which balances the training sets based on pseudo-labels obtained
from learned representations in GCL. We then develop the node centrality based
PBS method to better preserve the intrinsic structure of graphs, by upweighting
the important nodes of the given graph. Extensive experiments on multiple
imbalanced graph datasets and imbalanced settings demonstrate the effectiveness
of our proposed framework, which significantly improves the performance of the
recent state-of-the-art GCL methods. Further experimental ablations and
analyses show that the ImGCL framework consistently improves the representation
quality of nodes in under-represented (tail) classes.
| no_new_dataset | 0.711481 |
2205.14443 | Shaoru Wang | Shaoru Wang, Jin Gao, Zeming Li, Xiaoqin Zhang, Weiming Hu | A Closer Look at Self-Supervised Lightweight Vision Transformers | Accepted by ICML 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Self-supervised learning on large-scale Vision Transformers (ViTs) as
pre-training methods has achieved promising downstream performance. Yet, how
much these pre-training paradigms promote lightweight ViTs' performance is
considerably less studied. In this work, we develop and benchmark several
self-supervised pre-training methods on image classification tasks and some
downstream dense prediction tasks. We surprisingly find that if proper
pre-training is adopted, even vanilla lightweight ViTs show comparable
performance to previous SOTA networks with delicate architecture design. It
breaks the recently popular conception that vanilla ViTs are not suitable for
vision tasks in lightweight regimes. We also point out some defects of such
pre-training, e.g., failing to benefit from large-scale pre-training data and
showing inferior performance on data-insufficient downstream tasks.
Furthermore, we analyze and clearly show the effect of such pre-training by
analyzing the properties of the layer representation and attention maps for
related models. Finally, based on the above analyses, a distillation strategy
during pre-training is developed, which leads to further downstream performance
improvement for MAE-based pre-training. Code is available at
https://github.com/wangsr126/mae-lite.
| [
{
"version": "v1",
"created": "Sat, 28 May 2022 14:14:57 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 15:07:01 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Wang",
"Shaoru",
""
],
[
"Gao",
"Jin",
""
],
[
"Li",
"Zeming",
""
],
[
"Zhang",
"Xiaoqin",
""
],
[
"Hu",
"Weiming",
""
]
] | TITLE: A Closer Look at Self-Supervised Lightweight Vision Transformers
ABSTRACT: Self-supervised learning on large-scale Vision Transformers (ViTs) as
pre-training methods has achieved promising downstream performance. Yet, how
much these pre-training paradigms promote lightweight ViTs' performance is
considerably less studied. In this work, we develop and benchmark several
self-supervised pre-training methods on image classification tasks and some
downstream dense prediction tasks. We surprisingly find that if proper
pre-training is adopted, even vanilla lightweight ViTs show comparable
performance to previous SOTA networks with delicate architecture design. It
breaks the recently popular conception that vanilla ViTs are not suitable for
vision tasks in lightweight regimes. We also point out some defects of such
pre-training, e.g., failing to benefit from large-scale pre-training data and
showing inferior performance on data-insufficient downstream tasks.
Furthermore, we analyze and clearly show the effect of such pre-training by
analyzing the properties of the layer representation and attention maps for
related models. Finally, based on the above analyses, a distillation strategy
during pre-training is developed, which leads to further downstream performance
improvement for MAE-based pre-training. Code is available at
https://github.com/wangsr126/mae-lite.
| no_new_dataset | 0.709252 |
2206.00137 | Yiqiao Liao | Yiqiao Liao, Parinaz Naghizadeh | Social Bias Meets Data Bias: The Impacts of Labeling and Measurement
Errors on Fairness Criteria | null | null | null | null | cs.LG cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Although many fairness criteria have been proposed to ensure that machine
learning algorithms do not exhibit or amplify our existing social biases, these
algorithms are trained on datasets that can themselves be statistically biased.
In this paper, we investigate the robustness of a number of existing
(demographic) fairness criteria when the algorithm is trained on biased data.
We consider two forms of dataset bias: errors by prior decision makers in the
labeling process, and errors in measurement of the features of disadvantaged
individuals. We analytically show that some constraints (such as Demographic
Parity) can remain robust when facing certain statistical biases, while others
(such as Equalized Odds) are significantly violated if trained on biased data.
We also analyze the sensitivity of these criteria and the decision maker's
utility to biases. We provide numerical experiments based on three real-world
datasets (the FICO, Adult, and German credit score datasets) supporting our
analytical findings. Our findings present an additional guideline for choosing
among existing fairness criteria, or for proposing new criteria, when available
datasets may be biased.
| [
{
"version": "v1",
"created": "Tue, 31 May 2022 22:43:09 GMT"
},
{
"version": "v2",
"created": "Sun, 5 Mar 2023 08:27:32 GMT"
},
{
"version": "v3",
"created": "Mon, 10 Apr 2023 00:21:31 GMT"
},
{
"version": "v4",
"created": "Tue, 2 May 2023 18:10:10 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Liao",
"Yiqiao",
""
],
[
"Naghizadeh",
"Parinaz",
""
]
] | TITLE: Social Bias Meets Data Bias: The Impacts of Labeling and Measurement
Errors on Fairness Criteria
ABSTRACT: Although many fairness criteria have been proposed to ensure that machine
learning algorithms do not exhibit or amplify our existing social biases, these
algorithms are trained on datasets that can themselves be statistically biased.
In this paper, we investigate the robustness of a number of existing
(demographic) fairness criteria when the algorithm is trained on biased data.
We consider two forms of dataset bias: errors by prior decision makers in the
labeling process, and errors in measurement of the features of disadvantaged
individuals. We analytically show that some constraints (such as Demographic
Parity) can remain robust when facing certain statistical biases, while others
(such as Equalized Odds) are significantly violated if trained on biased data.
We also analyze the sensitivity of these criteria and the decision maker's
utility to biases. We provide numerical experiments based on three real-world
datasets (the FICO, Adult, and German credit score datasets) supporting our
analytical findings. Our findings present an additional guideline for choosing
among existing fairness criteria, or for proposing new criteria, when available
datasets may be biased.
| no_new_dataset | 0.713631 |
2206.08612 | Firat Ozdemir | Firat Ozdemir, Berkan Lafci, Xos\'e Lu\'is De\'an-Ben, Daniel
Razansky, Fernando Perez-Cruz | OADAT: Experimental and Synthetic Clinical Optoacoustic Data for
Standardized Image Processing | Accepted to TMLR. 32 pages, 24 figures, 9 tables | Transactions on Machine Learning Research (2023) 2835-8856 | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Optoacoustic (OA) imaging is based on excitation of biological tissues with
nanosecond-duration laser pulses followed by subsequent detection of ultrasound
waves generated via light-absorption-mediated thermoelastic expansion. OA
imaging features a powerful combination between rich optical contrast and high
resolution in deep tissues. This enabled the exploration of a number of
attractive new applications both in clinical and laboratory settings. However,
no standardized datasets generated with different types of experimental set-up
and associated processing methods are available to facilitate advances in
broader applications of OA in clinical settings. This complicates an objective
comparison between new and established data processing methods, often leading
to qualitative results and arbitrary interpretations of the data. In this
paper, we provide both experimental and synthetic OA raw signals and
reconstructed image domain datasets rendered with different experimental
parameters and tomographic acquisition geometries. We further provide trained
neural networks to tackle three important challenges related to OA image
processing, namely accurate reconstruction under limited view tomographic
conditions, removal of spatial undersampling artifacts and anatomical
segmentation for improved image reconstruction. Specifically, we define 44
experiments corresponding to the aforementioned challenges as benchmarks to be
used as a reference for the development of more advanced processing methods.
| [
{
"version": "v1",
"created": "Fri, 17 Jun 2022 08:11:26 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 15:40:04 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Ozdemir",
"Firat",
""
],
[
"Lafci",
"Berkan",
""
],
[
"Deán-Ben",
"Xosé Luís",
""
],
[
"Razansky",
"Daniel",
""
],
[
"Perez-Cruz",
"Fernando",
""
]
] | TITLE: OADAT: Experimental and Synthetic Clinical Optoacoustic Data for
Standardized Image Processing
ABSTRACT: Optoacoustic (OA) imaging is based on excitation of biological tissues with
nanosecond-duration laser pulses followed by subsequent detection of ultrasound
waves generated via light-absorption-mediated thermoelastic expansion. OA
imaging features a powerful combination between rich optical contrast and high
resolution in deep tissues. This enabled the exploration of a number of
attractive new applications both in clinical and laboratory settings. However,
no standardized datasets generated with different types of experimental set-up
and associated processing methods are available to facilitate advances in
broader applications of OA in clinical settings. This complicates an objective
comparison between new and established data processing methods, often leading
to qualitative results and arbitrary interpretations of the data. In this
paper, we provide both experimental and synthetic OA raw signals and
reconstructed image domain datasets rendered with different experimental
parameters and tomographic acquisition geometries. We further provide trained
neural networks to tackle three important challenges related to OA image
processing, namely accurate reconstruction under limited view tomographic
conditions, removal of spatial undersampling artifacts and anatomical
segmentation for improved image reconstruction. Specifically, we define 44
experiments corresponding to the aforementioned challenges as benchmarks to be
used as a reference for the development of more advanced processing methods.
| no_new_dataset | 0.712614 |
2207.12127 | Patrick Diehl | Nanmiao Wu and Ioannis Gonidelis and Simeng Liu and Zane Fink and
Nikunj Gupta and Karame Mohammadiporshokooh and Patrick Diehl and Hartmut
Kaiser and Laxmikant V. Kale | Quantifying Overheads in Charm++ and HPX using Task Bench | null | null | 10.1007/978-3-031-31209-0_1 | null | cs.DC cs.PF | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Asynchronous Many-Task (AMT) runtime systems take advantage of multi-core
architectures with light-weight threads, asynchronous executions, and smart
scheduling. In this paper, we present the comparison of the AMT systems Charm++
and HPX with the main stream MPI, OpenMP, and MPI+OpenMP libraries using the
Task Bench benchmarks. Charm++ is a parallel programming language based on C++,
supporting stackless tasks as well as light-weight threads asynchronously along
with an adaptive runtime system. HPX is a C++ library for concurrency and
parallelism, exposing C++ standards conforming API. First, we analyze the
commonalities, differences, and advantageous scenarios of Charm++ and HPX in
detail. Further, to investigate the potential overheads introduced by the
tasking systems of Charm++ and HPX, we utilize an existing parameterized
benchmark, Task Bench, wherein 15 different programming systems were
implemented, e.g., MPI, OpenMP, MPI + OpenMP, and extend Task Bench by adding
HPX implementations. We quantify the overheads of Charm++, HPX, and the main
stream libraries in different scenarios where a single task and multi-task are
assigned to each core, respectively. We also investigate each system's
scalability and the ability to hide the communication latency.
| [
{
"version": "v1",
"created": "Thu, 21 Jul 2022 18:33:01 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Wu",
"Nanmiao",
""
],
[
"Gonidelis",
"Ioannis",
""
],
[
"Liu",
"Simeng",
""
],
[
"Fink",
"Zane",
""
],
[
"Gupta",
"Nikunj",
""
],
[
"Mohammadiporshokooh",
"Karame",
""
],
[
"Diehl",
"Patrick",
""
],
[
"Kaiser",
"Hartmut",
""
],
[
"Kale",
"Laxmikant V.",
""
]
] | TITLE: Quantifying Overheads in Charm++ and HPX using Task Bench
ABSTRACT: Asynchronous Many-Task (AMT) runtime systems take advantage of multi-core
architectures with light-weight threads, asynchronous executions, and smart
scheduling. In this paper, we present the comparison of the AMT systems Charm++
and HPX with the main stream MPI, OpenMP, and MPI+OpenMP libraries using the
Task Bench benchmarks. Charm++ is a parallel programming language based on C++,
supporting stackless tasks as well as light-weight threads asynchronously along
with an adaptive runtime system. HPX is a C++ library for concurrency and
parallelism, exposing C++ standards conforming API. First, we analyze the
commonalities, differences, and advantageous scenarios of Charm++ and HPX in
detail. Further, to investigate the potential overheads introduced by the
tasking systems of Charm++ and HPX, we utilize an existing parameterized
benchmark, Task Bench, wherein 15 different programming systems were
implemented, e.g., MPI, OpenMP, MPI + OpenMP, and extend Task Bench by adding
HPX implementations. We quantify the overheads of Charm++, HPX, and the main
stream libraries in different scenarios where a single task and multi-task are
assigned to each core, respectively. We also investigate each system's
scalability and the ability to hide the communication latency.
| no_new_dataset | 0.708572 |
2209.03878 | Joshua Peeples | Joshua Peeples, Alina Zare, Jeffrey Dale, James Keller | Histogram Layers for Synthetic Aperture Sonar Imagery | 7 pages, 9 Figures, Accepted to IEEE International Conference on
Machine Learning and Applications (ICMLA) 2022 | null | 10.1109/ICMLA55696.2022.00032 | null | cs.CV cs.AI cs.LG eess.IV | http://creativecommons.org/licenses/by/4.0/ | Synthetic aperture sonar (SAS) imagery is crucial for several applications,
including target recognition and environmental segmentation. Deep learning
models have led to much success in SAS analysis; however, the features
extracted by these approaches may not be suitable for capturing certain
textural information. To address this problem, we present a novel application
of histogram layers on SAS imagery. The addition of histogram layer(s) within
the deep learning models improved performance by incorporating statistical
texture information on both synthetic and real-world datasets.
| [
{
"version": "v1",
"created": "Thu, 8 Sep 2022 15:33:35 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Peeples",
"Joshua",
""
],
[
"Zare",
"Alina",
""
],
[
"Dale",
"Jeffrey",
""
],
[
"Keller",
"James",
""
]
] | TITLE: Histogram Layers for Synthetic Aperture Sonar Imagery
ABSTRACT: Synthetic aperture sonar (SAS) imagery is crucial for several applications,
including target recognition and environmental segmentation. Deep learning
models have led to much success in SAS analysis; however, the features
extracted by these approaches may not be suitable for capturing certain
textural information. To address this problem, we present a novel application
of histogram layers on SAS imagery. The addition of histogram layer(s) within
the deep learning models improved performance by incorporating statistical
texture information on both synthetic and real-world datasets.
| no_new_dataset | 0.713032 |
2209.13052 | Nina Wiedemann | Nina Wiedemann, Valentin W\"uest, Antonio Loquercio, Matthias
M\"uller, Dario Floreano, Davide Scaramuzza | Training Efficient Controllers via Analytic Policy Gradient | null | IEEE Conference on Robotics and Automation (ICRA 2023) | null | null | cs.RO cs.AI | http://creativecommons.org/licenses/by/4.0/ | Control design for robotic systems is complex and often requires solving an
optimization to follow a trajectory accurately. Online optimization approaches
like Model Predictive Control (MPC) have been shown to achieve great tracking
performance, but require high computing power. Conversely, learning-based
offline optimization approaches, such as Reinforcement Learning (RL), allow
fast and efficient execution on the robot but hardly match the accuracy of MPC
in trajectory tracking tasks. In systems with limited compute, such as aerial
vehicles, an accurate controller that is efficient at execution time is
imperative. We propose an Analytic Policy Gradient (APG) method to tackle this
problem. APG exploits the availability of differentiable simulators by training
a controller offline with gradient descent on the tracking error. We address
training instabilities that frequently occur with APG through curriculum
learning and experiment on a widely used controls benchmark, the CartPole, and
two common aerial robots, a quadrotor and a fixed-wing drone. Our proposed
method outperforms both model-based and model-free RL methods in terms of
tracking error. Concurrently, it achieves similar performance to MPC while
requiring more than an order of magnitude less computation time. Our work
provides insights into the potential of APG as a promising control method for
robotics. To facilitate the exploration of APG, we open-source our code and
make it available at https://github.com/lis-epfl/apg_trajectory_tracking.
| [
{
"version": "v1",
"created": "Mon, 26 Sep 2022 22:04:35 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 18:14:44 GMT"
},
{
"version": "v3",
"created": "Tue, 2 May 2023 21:29:49 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Wiedemann",
"Nina",
""
],
[
"Wüest",
"Valentin",
""
],
[
"Loquercio",
"Antonio",
""
],
[
"Müller",
"Matthias",
""
],
[
"Floreano",
"Dario",
""
],
[
"Scaramuzza",
"Davide",
""
]
] | TITLE: Training Efficient Controllers via Analytic Policy Gradient
ABSTRACT: Control design for robotic systems is complex and often requires solving an
optimization to follow a trajectory accurately. Online optimization approaches
like Model Predictive Control (MPC) have been shown to achieve great tracking
performance, but require high computing power. Conversely, learning-based
offline optimization approaches, such as Reinforcement Learning (RL), allow
fast and efficient execution on the robot but hardly match the accuracy of MPC
in trajectory tracking tasks. In systems with limited compute, such as aerial
vehicles, an accurate controller that is efficient at execution time is
imperative. We propose an Analytic Policy Gradient (APG) method to tackle this
problem. APG exploits the availability of differentiable simulators by training
a controller offline with gradient descent on the tracking error. We address
training instabilities that frequently occur with APG through curriculum
learning and experiment on a widely used controls benchmark, the CartPole, and
two common aerial robots, a quadrotor and a fixed-wing drone. Our proposed
method outperforms both model-based and model-free RL methods in terms of
tracking error. Concurrently, it achieves similar performance to MPC while
requiring more than an order of magnitude less computation time. Our work
provides insights into the potential of APG as a promising control method for
robotics. To facilitate the exploration of APG, we open-source our code and
make it available at https://github.com/lis-epfl/apg_trajectory_tracking.
| no_new_dataset | 0.708036 |
2209.14488 | Gang Chen | Gang Chen and Victoria Huang | Ensemble Reinforcement Learning in Continuous Spaces -- A Hierarchical
Multi-Step Approach for Policy Training | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Actor-critic deep reinforcement learning (DRL) algorithms have recently
achieved prominent success in tackling various challenging reinforcement
learning (RL) problems, particularly complex control tasks with
high-dimensional continuous state and action spaces. Nevertheless, existing
research showed that actor-critic DRL algorithms often failed to explore their
learning environments effectively, resulting in limited learning stability and
performance. To address this limitation, several ensemble DRL algorithms have
been proposed lately to boost exploration and stabilize the learning process.
However, most of existing ensemble algorithms do not explicitly train all base
learners towards jointly optimizing the performance of the ensemble. In this
paper, we propose a new technique to train an ensemble of base learners based
on an innovative multi-step integration method. This training technique enables
us to develop a new hierarchical learning algorithm for ensemble DRL that
effectively promotes inter-learner collaboration through stable inter-learner
parameter sharing. The design of our new algorithm is verified theoretically.
The algorithm is also shown empirically to outperform several state-of-the-art
DRL algorithms on multiple benchmark RL problems.
| [
{
"version": "v1",
"created": "Thu, 29 Sep 2022 00:42:44 GMT"
},
{
"version": "v2",
"created": "Tue, 2 May 2023 23:44:10 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Chen",
"Gang",
""
],
[
"Huang",
"Victoria",
""
]
] | TITLE: Ensemble Reinforcement Learning in Continuous Spaces -- A Hierarchical
Multi-Step Approach for Policy Training
ABSTRACT: Actor-critic deep reinforcement learning (DRL) algorithms have recently
achieved prominent success in tackling various challenging reinforcement
learning (RL) problems, particularly complex control tasks with
high-dimensional continuous state and action spaces. Nevertheless, existing
research showed that actor-critic DRL algorithms often failed to explore their
learning environments effectively, resulting in limited learning stability and
performance. To address this limitation, several ensemble DRL algorithms have
been proposed lately to boost exploration and stabilize the learning process.
However, most of existing ensemble algorithms do not explicitly train all base
learners towards jointly optimizing the performance of the ensemble. In this
paper, we propose a new technique to train an ensemble of base learners based
on an innovative multi-step integration method. This training technique enables
us to develop a new hierarchical learning algorithm for ensemble DRL that
effectively promotes inter-learner collaboration through stable inter-learner
parameter sharing. The design of our new algorithm is verified theoretically.
The algorithm is also shown empirically to outperform several state-of-the-art
DRL algorithms on multiple benchmark RL problems.
| no_new_dataset | 0.709025 |
2210.01185 | Dejiao Zhang | Nihal Jain, Dejiao Zhang, Wasi Uddin Ahmad, Zijian Wang, Feng Nan,
Xiaopeng Li, Ming Tan, Ramesh Nallapati, Baishakhi Ray, Parminder Bhatia,
Xiaofei Ma, Bing Xiang | ContraCLM: Contrastive Learning For Causal Language Model | 10 pages | ACL 2023 | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Despite exciting progress in causal language models, the expressiveness of
the representations is largely limited due to poor discrimination ability. To
remedy this issue, we present ContraCLM, a novel contrastive learning framework
at both token-level and sequence-level. We assess ContraCLM on a variety of
downstream tasks. We show that ContraCLM enhances discrimination of the
representations and bridges the gap with the encoder-only models, which makes
causal language models better suited for tasks beyond language generation.
Specifically, we attain $44\%$ relative improvement on the Semantic Textual
Similarity tasks and $34\%$ on Code-to-Code Search tasks. Furthermore, by
improving the expressiveness of the representations, ContraCLM also boosts the
source code generation capability with $9\%$ relative improvement on execution
accuracy on the HumanEval benchmark.
| [
{
"version": "v1",
"created": "Mon, 3 Oct 2022 18:56:35 GMT"
},
{
"version": "v2",
"created": "Tue, 2 May 2023 22:46:46 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Jain",
"Nihal",
""
],
[
"Zhang",
"Dejiao",
""
],
[
"Ahmad",
"Wasi Uddin",
""
],
[
"Wang",
"Zijian",
""
],
[
"Nan",
"Feng",
""
],
[
"Li",
"Xiaopeng",
""
],
[
"Tan",
"Ming",
""
],
[
"Nallapati",
"Ramesh",
""
],
[
"Ray",
"Baishakhi",
""
],
[
"Bhatia",
"Parminder",
""
],
[
"Ma",
"Xiaofei",
""
],
[
"Xiang",
"Bing",
""
]
] | TITLE: ContraCLM: Contrastive Learning For Causal Language Model
ABSTRACT: Despite exciting progress in causal language models, the expressiveness of
the representations is largely limited due to poor discrimination ability. To
remedy this issue, we present ContraCLM, a novel contrastive learning framework
at both token-level and sequence-level. We assess ContraCLM on a variety of
downstream tasks. We show that ContraCLM enhances discrimination of the
representations and bridges the gap with the encoder-only models, which makes
causal language models better suited for tasks beyond language generation.
Specifically, we attain $44\%$ relative improvement on the Semantic Textual
Similarity tasks and $34\%$ on Code-to-Code Search tasks. Furthermore, by
improving the expressiveness of the representations, ContraCLM also boosts the
source code generation capability with $9\%$ relative improvement on execution
accuracy on the HumanEval benchmark.
| no_new_dataset | 0.710409 |
2210.05311 | Wuti Xiong | Wuti Xiong | CD-FSOD: A Benchmark for Cross-domain Few-shot Object Detection | Accepted by ICASSP 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a study of the cross-domain few-shot object
detection (CD-FSOD) benchmark, consisting of image data from a diverse data
domain. On the proposed benchmark, we evaluate state-of-art FSOD approaches,
including meta-learning FSOD approaches and fine-tuning FSOD approaches. The
results show that these methods tend to fall, and even underperform the naive
fine-tuning model. We analyze the reasons for their failure and introduce a
strong baseline that uses a mutually-beneficial manner to alleviate the
overfitting problem. Our approach is remarkably superior to existing approaches
by significant margins (2.0\% on average) on the proposed benchmark. Our code
is available at \url{https://github.com/FSOD/CD-FSOD}.
| [
{
"version": "v1",
"created": "Tue, 11 Oct 2022 10:10:07 GMT"
},
{
"version": "v2",
"created": "Thu, 27 Oct 2022 10:52:27 GMT"
},
{
"version": "v3",
"created": "Wed, 3 May 2023 09:19:05 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Xiong",
"Wuti",
""
]
] | TITLE: CD-FSOD: A Benchmark for Cross-domain Few-shot Object Detection
ABSTRACT: In this paper, we propose a study of the cross-domain few-shot object
detection (CD-FSOD) benchmark, consisting of image data from a diverse data
domain. On the proposed benchmark, we evaluate state-of-art FSOD approaches,
including meta-learning FSOD approaches and fine-tuning FSOD approaches. The
results show that these methods tend to fall, and even underperform the naive
fine-tuning model. We analyze the reasons for their failure and introduce a
strong baseline that uses a mutually-beneficial manner to alleviate the
overfitting problem. Our approach is remarkably superior to existing approaches
by significant margins (2.0\% on average) on the proposed benchmark. Our code
is available at \url{https://github.com/FSOD/CD-FSOD}.
| no_new_dataset | 0.706437 |
2210.08360 | Parker Lusk | Parker C. Lusk, Kaveh Fathian, Jonathan P. How | MIXER: Multiattribute, Multiway Fusion of Uncertain Pairwise Affinities | 8 pages + proofs | IEEE Robotics and Automation Letters, vol. 8, no. 5, pp.
2462-2469, May 2023 | 10.1109/LRA.2023.3253025 | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a multiway fusion algorithm capable of directly processing
uncertain pairwise affinities. In contrast to existing works that require
initial pairwise associations, our MIXER algorithm improves accuracy by
leveraging the additional information provided by pairwise affinities. Our main
contribution is a multiway fusion formulation that is particularly suited to
processing non-binary affinities and a novel continuous relaxation whose
solutions are guaranteed to be binary, thus avoiding the typical, but
potentially problematic, solution binarization steps that may cause
infeasibility. A crucial insight of our formulation is that it allows for three
modes of association, ranging from non-match, undecided, and match. Exploiting
this insight allows fusion to be delayed for some data pairs until more
information is available, which is an effective feature for fusion of data with
multiple attributes/information sources. We evaluate MIXER on typical synthetic
data and benchmark datasets and show increased accuracy against the state of
the art in multiway matching, especially in noisy regimes with low observation
redundancy. Additionally, we collect RGB data of cars in a parking lot to
demonstrate MIXER's ability to fuse data having multiple attributes (color,
visual appearance, and bounding box). On this challenging dataset, MIXER
achieves 74% F1 accuracy and is 49x faster than the next best algorithm, which
has 42% accuracy. Open source code is available at
https://github.com/mit-acl/mixer.
| [
{
"version": "v1",
"created": "Sat, 15 Oct 2022 19:19:57 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 15:28:43 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Lusk",
"Parker C.",
""
],
[
"Fathian",
"Kaveh",
""
],
[
"How",
"Jonathan P.",
""
]
] | TITLE: MIXER: Multiattribute, Multiway Fusion of Uncertain Pairwise Affinities
ABSTRACT: We present a multiway fusion algorithm capable of directly processing
uncertain pairwise affinities. In contrast to existing works that require
initial pairwise associations, our MIXER algorithm improves accuracy by
leveraging the additional information provided by pairwise affinities. Our main
contribution is a multiway fusion formulation that is particularly suited to
processing non-binary affinities and a novel continuous relaxation whose
solutions are guaranteed to be binary, thus avoiding the typical, but
potentially problematic, solution binarization steps that may cause
infeasibility. A crucial insight of our formulation is that it allows for three
modes of association, ranging from non-match, undecided, and match. Exploiting
this insight allows fusion to be delayed for some data pairs until more
information is available, which is an effective feature for fusion of data with
multiple attributes/information sources. We evaluate MIXER on typical synthetic
data and benchmark datasets and show increased accuracy against the state of
the art in multiway matching, especially in noisy regimes with low observation
redundancy. Additionally, we collect RGB data of cars in a parking lot to
demonstrate MIXER's ability to fuse data having multiple attributes (color,
visual appearance, and bounding box). On this challenging dataset, MIXER
achieves 74% F1 accuracy and is 49x faster than the next best algorithm, which
has 42% accuracy. Open source code is available at
https://github.com/mit-acl/mixer.
| no_new_dataset | 0.708648 |
2210.08682 | Tingyuan Liang | Tingyuan Liang, Gengjie Chen, Jieru Zhao, Sharad Sinha, Wei Zhang | AMF-Placer 2.0: Open Source Timing-driven Analytical Mixed-size Placer
for Large-scale Heterogeneous FPGA | null | null | null | null | cs.AR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | On modern field-programmable gate arrays (FPGAs), certain critical path
portions of the designs might be prearranged into many multi-cell macros during
synthesis. These movable macros with constraints of shape and resources lead to
challenging mixed-size placement for FPGA designs which cannot be addressed by
previous analytical placers. Moreover, general timing-driven placement
algorithms are facing challenges when handling real-world application design
and ultrascale FPGA architectures. In this work, we propose AMF-Placer 2.0, an
open-source comprehensive timing-driven analytical mixed-size FPGA placer. It
supports mixed-size placement of heterogeneous resources (e.g.,
LUT/FF/LUTRAM/MUX/CARRY/DSP/BRAM) on FPGA, with an interface to Xilinx Vivado.
Standing upon the shoulders of AMF-Placer 1.0, AMFPlacer 2.0 is equipped with a
series of new techniques for timing optimization, including a simple but
effective timing model, placement-blockage-aware anchor insertion, WNS-aware
timing-driven quadratic placement, and sector-guided detailed placement. Based
on a set of the latest large open-source benchmarks from various domains for
Xilinx Ultrascale FPGAs, experimental results indicate that critical path
delays realized by AMF-Placer 2.0 are averagely 2.2% and 0.59% higher than
those achieved by commercial tool Xilinx Vivavo 2020.2 and 2021.2 respectively.
Meanwhile, the average runtime of placement procedure of AMF-Placer 2.0 is 14%
and 8.5% higher than Xilinx Vivavo 2020.2 and 2021.2 respectively. Although
limited by the absence of the exact timing model of the device, the information
of design hierarchy and accurate routing feedback, AMF-Placer 2.0 is the first
open-source FPGA placer which can handle the timingdriven mixed-size placement
of practical complex designs with various FPGA resources and achieves the
comparable quality to the latest commercial tools.
| [
{
"version": "v1",
"created": "Mon, 17 Oct 2022 01:04:21 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 04:57:04 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Liang",
"Tingyuan",
""
],
[
"Chen",
"Gengjie",
""
],
[
"Zhao",
"Jieru",
""
],
[
"Sinha",
"Sharad",
""
],
[
"Zhang",
"Wei",
""
]
] | TITLE: AMF-Placer 2.0: Open Source Timing-driven Analytical Mixed-size Placer
for Large-scale Heterogeneous FPGA
ABSTRACT: On modern field-programmable gate arrays (FPGAs), certain critical path
portions of the designs might be prearranged into many multi-cell macros during
synthesis. These movable macros with constraints of shape and resources lead to
challenging mixed-size placement for FPGA designs which cannot be addressed by
previous analytical placers. Moreover, general timing-driven placement
algorithms are facing challenges when handling real-world application design
and ultrascale FPGA architectures. In this work, we propose AMF-Placer 2.0, an
open-source comprehensive timing-driven analytical mixed-size FPGA placer. It
supports mixed-size placement of heterogeneous resources (e.g.,
LUT/FF/LUTRAM/MUX/CARRY/DSP/BRAM) on FPGA, with an interface to Xilinx Vivado.
Standing upon the shoulders of AMF-Placer 1.0, AMFPlacer 2.0 is equipped with a
series of new techniques for timing optimization, including a simple but
effective timing model, placement-blockage-aware anchor insertion, WNS-aware
timing-driven quadratic placement, and sector-guided detailed placement. Based
on a set of the latest large open-source benchmarks from various domains for
Xilinx Ultrascale FPGAs, experimental results indicate that critical path
delays realized by AMF-Placer 2.0 are averagely 2.2% and 0.59% higher than
those achieved by commercial tool Xilinx Vivavo 2020.2 and 2021.2 respectively.
Meanwhile, the average runtime of placement procedure of AMF-Placer 2.0 is 14%
and 8.5% higher than Xilinx Vivavo 2020.2 and 2021.2 respectively. Although
limited by the absence of the exact timing model of the device, the information
of design hierarchy and accurate routing feedback, AMF-Placer 2.0 is the first
open-source FPGA placer which can handle the timingdriven mixed-size placement
of practical complex designs with various FPGA resources and achieves the
comparable quality to the latest commercial tools.
| no_new_dataset | 0.710377 |
2211.01317 | Huck Yang | Yun-Ning Hung, Chao-Han Huck Yang, Pin-Yu Chen, Alexander Lerch | Low-Resource Music Genre Classification with Cross-Modal Neural Model
Reprogramming | Accepted to IEEE ICASSP 2023. The implementation is available at
https://github.com/biboamy/music-repro | null | null | null | cs.SD cs.AI cs.LG cs.NE eess.AS | http://creativecommons.org/licenses/by/4.0/ | Transfer learning (TL) approaches have shown promising results when handling
tasks with limited training data. However, considerable memory and
computational resources are often required for fine-tuning pre-trained neural
networks with target domain data. In this work, we introduce a novel method for
leveraging pre-trained models for low-resource (music) classification based on
the concept of Neural Model Reprogramming (NMR). NMR aims at re-purposing a
pre-trained model from a source domain to a target domain by modifying the
input of a frozen pre-trained model. In addition to the known,
input-independent, reprogramming method, we propose an advanced reprogramming
paradigm: Input-dependent NMR, to increase adaptability to complex input data
such as musical audio. Experimental results suggest that a neural model
pre-trained on large-scale datasets can successfully perform music genre
classification by using this reprogramming method. The two proposed
Input-dependent NMR TL methods outperform fine-tuning-based TL methods on a
small genre classification dataset.
| [
{
"version": "v1",
"created": "Wed, 2 Nov 2022 17:38:33 GMT"
},
{
"version": "v2",
"created": "Fri, 28 Apr 2023 05:19:07 GMT"
},
{
"version": "v3",
"created": "Wed, 3 May 2023 04:22:54 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Hung",
"Yun-Ning",
""
],
[
"Yang",
"Chao-Han Huck",
""
],
[
"Chen",
"Pin-Yu",
""
],
[
"Lerch",
"Alexander",
""
]
] | TITLE: Low-Resource Music Genre Classification with Cross-Modal Neural Model
Reprogramming
ABSTRACT: Transfer learning (TL) approaches have shown promising results when handling
tasks with limited training data. However, considerable memory and
computational resources are often required for fine-tuning pre-trained neural
networks with target domain data. In this work, we introduce a novel method for
leveraging pre-trained models for low-resource (music) classification based on
the concept of Neural Model Reprogramming (NMR). NMR aims at re-purposing a
pre-trained model from a source domain to a target domain by modifying the
input of a frozen pre-trained model. In addition to the known,
input-independent, reprogramming method, we propose an advanced reprogramming
paradigm: Input-dependent NMR, to increase adaptability to complex input data
such as musical audio. Experimental results suggest that a neural model
pre-trained on large-scale datasets can successfully perform music genre
classification by using this reprogramming method. The two proposed
Input-dependent NMR TL methods outperform fine-tuning-based TL methods on a
small genre classification dataset.
| no_new_dataset | 0.711281 |
2211.06547 | Sandeep Reddy Kothinti | Sandeep Kothinti and Dimitra Emmanouilidou | Investigations in Audio Captioning: Addressing Vocabulary Imbalance and
Evaluating Suitability of Language-Centric Performance Metrics | Submitted to EUSIPCO 2023 | null | null | null | eess.AS cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The analysis, processing, and extraction of meaningful information from
sounds all around us is the subject of the broader area of audio analytics.
Audio captioning is a recent addition to the domain of audio analytics, a
cross-modal translation task that focuses on generating natural descriptions
from sound events occurring in an audio stream. In this work, we identify and
improve on three main challenges in automated audio captioning: i) data
scarcity, ii) imbalance or limitations in the audio captions vocabulary, and
iii) the proper performance evaluation metric that can best capture both
auditory and semantic characteristics. We find that generally adopted loss
functions can result in an unfair vocabulary imbalance during model training.
We propose two audio captioning augmentation methods that enrich the training
dataset and the vocabulary size. We further underline the need for in-domain
pretraining by exploring the suitability of audio encoders that were previously
trained on different audio tasks. Finally, we systematically explore five
performance metrics borrowed from the image captioning domain and highlight
their limitations for the audio domain.
| [
{
"version": "v1",
"created": "Sat, 12 Nov 2022 02:12:21 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 17:53:19 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Kothinti",
"Sandeep",
""
],
[
"Emmanouilidou",
"Dimitra",
""
]
] | TITLE: Investigations in Audio Captioning: Addressing Vocabulary Imbalance and
Evaluating Suitability of Language-Centric Performance Metrics
ABSTRACT: The analysis, processing, and extraction of meaningful information from
sounds all around us is the subject of the broader area of audio analytics.
Audio captioning is a recent addition to the domain of audio analytics, a
cross-modal translation task that focuses on generating natural descriptions
from sound events occurring in an audio stream. In this work, we identify and
improve on three main challenges in automated audio captioning: i) data
scarcity, ii) imbalance or limitations in the audio captions vocabulary, and
iii) the proper performance evaluation metric that can best capture both
auditory and semantic characteristics. We find that generally adopted loss
functions can result in an unfair vocabulary imbalance during model training.
We propose two audio captioning augmentation methods that enrich the training
dataset and the vocabulary size. We further underline the need for in-domain
pretraining by exploring the suitability of audio encoders that were previously
trained on different audio tasks. Finally, we systematically explore five
performance metrics borrowed from the image captioning domain and highlight
their limitations for the audio domain.
| no_new_dataset | 0.709617 |
2211.06761 | Umaid Muhammad Zaffar | Saad Bin Ahmed, Umaid M. Zaffar, Marium Aslam and Muhammad Imran Malik | Few-Shot Learning for Biometric Verification | 19 pages, 7 figures | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In machine learning applications, it is common practice to feed as much
information as possible. In most cases, the model can handle large data sets
that allow to predict more accurately. In the presence of data scarcity, a
Few-Shot learning (FSL) approach aims to build more accurate algorithms with
limited training data. We propose a novel end-to-end lightweight architecture
that verifies biometric data by producing competitive results as compared to
state-of-the-art accuracies through Few-Shot learning methods. The dense layers
add to the complexity of state-of-the-art deep learning models which inhibits
them to be used in low-power applications. In presented approach, a shallow
network is coupled with a conventional machine learning technique that exploits
hand-crafted features to verify biometric images from multi-modal sources such
as signatures, periocular region, iris, face, fingerprints etc. We introduce a
self-estimated threshold that strictly monitors False Acceptance Rate (FAR)
while generalizing its results hence eliminating user-defined thresholds from
ROC curves that are likely to be biased on local data distribution. This hybrid
model benefits from few-shot learning to make up for scarcity of data in
biometric use-cases. We have conducted extensive experimentation with commonly
used biometric datasets. The obtained results provided an effective solution
for biometric verification systems.
| [
{
"version": "v1",
"created": "Sat, 12 Nov 2022 22:49:25 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 15:24:02 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Ahmed",
"Saad Bin",
""
],
[
"Zaffar",
"Umaid M.",
""
],
[
"Aslam",
"Marium",
""
],
[
"Malik",
"Muhammad Imran",
""
]
] | TITLE: Few-Shot Learning for Biometric Verification
ABSTRACT: In machine learning applications, it is common practice to feed as much
information as possible. In most cases, the model can handle large data sets
that allow to predict more accurately. In the presence of data scarcity, a
Few-Shot learning (FSL) approach aims to build more accurate algorithms with
limited training data. We propose a novel end-to-end lightweight architecture
that verifies biometric data by producing competitive results as compared to
state-of-the-art accuracies through Few-Shot learning methods. The dense layers
add to the complexity of state-of-the-art deep learning models which inhibits
them to be used in low-power applications. In presented approach, a shallow
network is coupled with a conventional machine learning technique that exploits
hand-crafted features to verify biometric images from multi-modal sources such
as signatures, periocular region, iris, face, fingerprints etc. We introduce a
self-estimated threshold that strictly monitors False Acceptance Rate (FAR)
while generalizing its results hence eliminating user-defined thresholds from
ROC curves that are likely to be biased on local data distribution. This hybrid
model benefits from few-shot learning to make up for scarcity of data in
biometric use-cases. We have conducted extensive experimentation with commonly
used biometric datasets. The obtained results provided an effective solution
for biometric verification systems.
| no_new_dataset | 0.711462 |
2212.00419 | Francesco Pierri | Francesco Pierri, Luca Luceri, Nikhil Jindal, Emilio Ferrara | Propaganda and Misinformation on Facebook and Twitter during the Russian
Invasion of Ukraine | Accepted at WebSci'2023 | WebSci '23: Proceedings of the 15th ACM Web Science Conference
2023, April 2023, Pages 65-74 | 10.1145/3578503.3583597 | null | cs.SI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Online social media represent an oftentimes unique source of information, and
having access to reliable and unbiased content is crucial, especially during
crises and contentious events. We study the spread of propaganda and
misinformation that circulated on Facebook and Twitter during the first few
months of the Russia-Ukraine conflict. By leveraging two large datasets of
millions of social media posts, we estimate the prevalence of Russian
propaganda and low-credibility content on the two platforms, describing
temporal patterns and highlighting the disproportionate role played by
superspreaders in amplifying unreliable content. We infer the political leaning
of Facebook pages and Twitter users sharing propaganda and misinformation, and
observe they tend to be more right-leaning than the average. By estimating the
amount of content moderated by the two platforms, we show that only about 8-15%
of the posts and tweets sharing links to Russian propaganda or untrustworthy
sources were removed. Overall, our findings show that Facebook and Twitter are
still vulnerable to abuse, especially during crises: we highlight the need to
urgently address this issue to preserve the integrity of online conversations.
| [
{
"version": "v1",
"created": "Thu, 1 Dec 2022 10:49:18 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Feb 2023 09:31:34 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Pierri",
"Francesco",
""
],
[
"Luceri",
"Luca",
""
],
[
"Jindal",
"Nikhil",
""
],
[
"Ferrara",
"Emilio",
""
]
] | TITLE: Propaganda and Misinformation on Facebook and Twitter during the Russian
Invasion of Ukraine
ABSTRACT: Online social media represent an oftentimes unique source of information, and
having access to reliable and unbiased content is crucial, especially during
crises and contentious events. We study the spread of propaganda and
misinformation that circulated on Facebook and Twitter during the first few
months of the Russia-Ukraine conflict. By leveraging two large datasets of
millions of social media posts, we estimate the prevalence of Russian
propaganda and low-credibility content on the two platforms, describing
temporal patterns and highlighting the disproportionate role played by
superspreaders in amplifying unreliable content. We infer the political leaning
of Facebook pages and Twitter users sharing propaganda and misinformation, and
observe they tend to be more right-leaning than the average. By estimating the
amount of content moderated by the two platforms, we show that only about 8-15%
of the posts and tweets sharing links to Russian propaganda or untrustworthy
sources were removed. Overall, our findings show that Facebook and Twitter are
still vulnerable to abuse, especially during crises: we highlight the need to
urgently address this issue to preserve the integrity of online conversations.
| no_new_dataset | 0.702951 |
2212.02421 | Tim Hsu | Tim Hsu, Babak Sadigh, Nicolas Bertin, Cheol Woo Park, James Chapman,
Vasily Bulatov, Fei Zhou | Score-based denoising for atomic structure identification | null | null | null | null | cond-mat.mtrl-sci cs.LG physics.atom-ph | http://creativecommons.org/licenses/by/4.0/ | We propose an effective method for removing thermal vibrations that
complicate the task of analyzing complex dynamics in atomistic simulation of
condensed matter. Our method iteratively subtracts thermal noises or
perturbations in atomic positions using a denoising score function trained on
synthetically noised but otherwise perfect crystal lattices. The resulting
denoised structures clearly reveal underlying crystal order while retaining
disorder associated with crystal defects. Purely geometric, agnostic to
interatomic potentials, and trained without inputs from explicit simulations,
our denoiser can be applied to simulation data generated from vastly different
interatomic interactions. The denoiser is shown to improve existing
classification methods such as common neighbor analysis and polyhedral template
matching, reaching perfect classification accuracy on a recent benchmark
dataset of thermally perturbed structures up to the melting point. Demonstrated
here in a wide variety of atomistic simulation contexts, the denoiser is
general, robust, and readily extendable to delineate order from disorder in
structurally and chemically complex materials.
| [
{
"version": "v1",
"created": "Mon, 5 Dec 2022 17:18:17 GMT"
},
{
"version": "v2",
"created": "Tue, 20 Dec 2022 22:32:59 GMT"
},
{
"version": "v3",
"created": "Wed, 3 May 2023 06:00:22 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Hsu",
"Tim",
""
],
[
"Sadigh",
"Babak",
""
],
[
"Bertin",
"Nicolas",
""
],
[
"Park",
"Cheol Woo",
""
],
[
"Chapman",
"James",
""
],
[
"Bulatov",
"Vasily",
""
],
[
"Zhou",
"Fei",
""
]
] | TITLE: Score-based denoising for atomic structure identification
ABSTRACT: We propose an effective method for removing thermal vibrations that
complicate the task of analyzing complex dynamics in atomistic simulation of
condensed matter. Our method iteratively subtracts thermal noises or
perturbations in atomic positions using a denoising score function trained on
synthetically noised but otherwise perfect crystal lattices. The resulting
denoised structures clearly reveal underlying crystal order while retaining
disorder associated with crystal defects. Purely geometric, agnostic to
interatomic potentials, and trained without inputs from explicit simulations,
our denoiser can be applied to simulation data generated from vastly different
interatomic interactions. The denoiser is shown to improve existing
classification methods such as common neighbor analysis and polyhedral template
matching, reaching perfect classification accuracy on a recent benchmark
dataset of thermally perturbed structures up to the melting point. Demonstrated
here in a wide variety of atomistic simulation contexts, the denoiser is
general, robust, and readily extendable to delineate order from disorder in
structurally and chemically complex materials.
| no_new_dataset | 0.663396 |
2212.02504 | Jonas Ditz | Jonas C. Ditz, Bernhard Reuter, Nico Pfeifer | COmic: Convolutional Kernel Networks for Interpretable End-to-End
Learning on (Multi-)Omics Data | null | null | null | null | q-bio.QM cs.LG | http://creativecommons.org/licenses/by/4.0/ | Motivation: The size of available omics datasets is steadily increasing with
technological advancement in recent years. While this increase in sample size
can be used to improve the performance of relevant prediction tasks in
healthcare, models that are optimized for large datasets usually operate as
black boxes. In high stakes scenarios, like healthcare, using a black-box model
poses safety and security issues. Without an explanation about molecular
factors and phenotypes that affected the prediction, healthcare providers are
left with no choice but to blindly trust the models. We propose a new type of
artificial neural network, named Convolutional Omics Kernel Network (COmic). By
combining convolutional kernel networks with pathway-induced kernels, our
method enables robust and interpretable end-to-end learning on omics datasets
ranging in size from a few hundred to several hundreds of thousands of samples.
Furthermore, COmic can be easily adapted to utilize multi-omics data.
Results: We evaluated the performance capabilities of COmic on six different
breast cancer cohorts. Additionally, we trained COmic models on multi-omics
data using the METABRIC cohort. Our models performed either better or similar
to competitors on both tasks. We show how the use of pathway-induced Laplacian
kernels opens the black-box nature of neural networks and results in
intrinsically interpretable models that eliminate the need for post-hoc
explanation models.
| [
{
"version": "v1",
"created": "Fri, 2 Dec 2022 12:54:27 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 11:51:34 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Ditz",
"Jonas C.",
""
],
[
"Reuter",
"Bernhard",
""
],
[
"Pfeifer",
"Nico",
""
]
] | TITLE: COmic: Convolutional Kernel Networks for Interpretable End-to-End
Learning on (Multi-)Omics Data
ABSTRACT: Motivation: The size of available omics datasets is steadily increasing with
technological advancement in recent years. While this increase in sample size
can be used to improve the performance of relevant prediction tasks in
healthcare, models that are optimized for large datasets usually operate as
black boxes. In high stakes scenarios, like healthcare, using a black-box model
poses safety and security issues. Without an explanation about molecular
factors and phenotypes that affected the prediction, healthcare providers are
left with no choice but to blindly trust the models. We propose a new type of
artificial neural network, named Convolutional Omics Kernel Network (COmic). By
combining convolutional kernel networks with pathway-induced kernels, our
method enables robust and interpretable end-to-end learning on omics datasets
ranging in size from a few hundred to several hundreds of thousands of samples.
Furthermore, COmic can be easily adapted to utilize multi-omics data.
Results: We evaluated the performance capabilities of COmic on six different
breast cancer cohorts. Additionally, we trained COmic models on multi-omics
data using the METABRIC cohort. Our models performed either better or similar
to competitors on both tasks. We show how the use of pathway-induced Laplacian
kernels opens the black-box nature of neural networks and results in
intrinsically interpretable models that eliminate the need for post-hoc
explanation models.
| no_new_dataset | 0.711224 |
2212.09736 | Yu Gu | Yu Gu, Xiang Deng, Yu Su | Don't Generate, Discriminate: A Proposal for Grounding Language Models
to Real-World Environments | 18 pages, 6 figures, 6 tables; accepted to ACL'2023 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | A key missing capacity of current language models (LMs) is grounding to
real-world environments. Most existing work for grounded language understanding
uses LMs to directly generate plans that can be executed in the environment to
achieve the desired effects. It thereby casts the burden of ensuring
grammaticality, faithfulness, and controllability all on the LMs. We propose
Pangu, a generic framework for grounded language understanding that capitalizes
on the discriminative ability of LMs instead of their generative ability. Pangu
consists of a symbolic agent and a neural LM working in a concerted fashion:
The agent explores the environment to incrementally construct valid plans, and
the LM evaluates the plausibility of the candidate plans to guide the search
process. A case study on the challenging problem of knowledge base question
answering (KBQA), which features a massive environment, demonstrates the
remarkable effectiveness and flexibility of Pangu: A BERT-base LM is sufficient
for setting a new record on standard KBQA datasets, and larger LMs further
bring substantial gains. Pangu also enables, for the first time, effective
few-shot in-context learning for KBQA with large LMs such as Codex.
| [
{
"version": "v1",
"created": "Mon, 19 Dec 2022 18:55:21 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 04:32:35 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Gu",
"Yu",
""
],
[
"Deng",
"Xiang",
""
],
[
"Su",
"Yu",
""
]
] | TITLE: Don't Generate, Discriminate: A Proposal for Grounding Language Models
to Real-World Environments
ABSTRACT: A key missing capacity of current language models (LMs) is grounding to
real-world environments. Most existing work for grounded language understanding
uses LMs to directly generate plans that can be executed in the environment to
achieve the desired effects. It thereby casts the burden of ensuring
grammaticality, faithfulness, and controllability all on the LMs. We propose
Pangu, a generic framework for grounded language understanding that capitalizes
on the discriminative ability of LMs instead of their generative ability. Pangu
consists of a symbolic agent and a neural LM working in a concerted fashion:
The agent explores the environment to incrementally construct valid plans, and
the LM evaluates the plausibility of the candidate plans to guide the search
process. A case study on the challenging problem of knowledge base question
answering (KBQA), which features a massive environment, demonstrates the
remarkable effectiveness and flexibility of Pangu: A BERT-base LM is sufficient
for setting a new record on standard KBQA datasets, and larger LMs further
bring substantial gains. Pangu also enables, for the first time, effective
few-shot in-context learning for KBQA with large LMs such as Codex.
| no_new_dataset | 0.710791 |
2212.10375 | Zhiyong Wu | Zhiyong Wu, Yaoxiang Wang, Jiacheng Ye, Lingpeng Kong | Self-Adaptive In-Context Learning: An Information Compression
Perspective for In-Context Example Selection and Ordering | To appear at ACL 2023 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Despite the surprising few-shot performance of in-context learning (ICL), it
is still a common practice to randomly sample examples to serve as context.
This paper advocates a new principle for ICL: self-adaptive in-context
learning. The self-adaption mechanism is introduced to help each sample find an
in-context example permutation (i.e., selection and ordering) that can derive
the correct prediction, thus maximizing performance. To validate the
effectiveness of self-adaptive ICL, we propose a general select-then-rank
framework and instantiate it with new selection and ranking algorithms. Upon
extensive evaluation on eight different NLP datasets, our self-adaptive ICL
method achieves a 40% relative improvement over the common practice setting.
Further analysis reveals the enormous potential of self-adaptive ICL that it
might be able to close the gap between ICL and finetuning given more advanced
algorithms. Our code is released to facilitate future research in this area:
https://github.com/Shark-NLP/self-adaptive-ICL
| [
{
"version": "v1",
"created": "Tue, 20 Dec 2022 15:55:21 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 14:43:50 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Wu",
"Zhiyong",
""
],
[
"Wang",
"Yaoxiang",
""
],
[
"Ye",
"Jiacheng",
""
],
[
"Kong",
"Lingpeng",
""
]
] | TITLE: Self-Adaptive In-Context Learning: An Information Compression
Perspective for In-Context Example Selection and Ordering
ABSTRACT: Despite the surprising few-shot performance of in-context learning (ICL), it
is still a common practice to randomly sample examples to serve as context.
This paper advocates a new principle for ICL: self-adaptive in-context
learning. The self-adaption mechanism is introduced to help each sample find an
in-context example permutation (i.e., selection and ordering) that can derive
the correct prediction, thus maximizing performance. To validate the
effectiveness of self-adaptive ICL, we propose a general select-then-rank
framework and instantiate it with new selection and ranking algorithms. Upon
extensive evaluation on eight different NLP datasets, our self-adaptive ICL
method achieves a 40% relative improvement over the common practice setting.
Further analysis reveals the enormous potential of self-adaptive ICL that it
might be able to close the gap between ICL and finetuning given more advanced
algorithms. Our code is released to facilitate future research in this area:
https://github.com/Shark-NLP/self-adaptive-ICL
| no_new_dataset | 0.707443 |
2212.10562 | Noah Constant | Rosanne Liu, Dan Garrette, Chitwan Saharia, William Chan, Adam
Roberts, Sharan Narang, Irina Blok, RJ Mical, Mohammad Norouzi, Noah Constant | Character-Aware Models Improve Visual Text Rendering | null | null | null | null | cs.CL cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Current image generation models struggle to reliably produce well-formed
visual text. In this paper, we investigate a key contributing factor: popular
text-to-image models lack character-level input features, making it much harder
to predict a word's visual makeup as a series of glyphs. To quantify this
effect, we conduct a series of experiments comparing character-aware vs.
character-blind text encoders. In the text-only domain, we find that
character-aware models provide large gains on a novel spelling task
(WikiSpell). Applying our learnings to the visual domain, we train a suite of
image generation models, and show that character-aware variants outperform
their character-blind counterparts across a range of novel text rendering tasks
(our DrawText benchmark). Our models set a much higher state-of-the-art on
visual spelling, with 30+ point accuracy gains over competitors on rare words,
despite training on far fewer examples.
| [
{
"version": "v1",
"created": "Tue, 20 Dec 2022 18:59:23 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 16:36:38 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Liu",
"Rosanne",
""
],
[
"Garrette",
"Dan",
""
],
[
"Saharia",
"Chitwan",
""
],
[
"Chan",
"William",
""
],
[
"Roberts",
"Adam",
""
],
[
"Narang",
"Sharan",
""
],
[
"Blok",
"Irina",
""
],
[
"Mical",
"RJ",
""
],
[
"Norouzi",
"Mohammad",
""
],
[
"Constant",
"Noah",
""
]
] | TITLE: Character-Aware Models Improve Visual Text Rendering
ABSTRACT: Current image generation models struggle to reliably produce well-formed
visual text. In this paper, we investigate a key contributing factor: popular
text-to-image models lack character-level input features, making it much harder
to predict a word's visual makeup as a series of glyphs. To quantify this
effect, we conduct a series of experiments comparing character-aware vs.
character-blind text encoders. In the text-only domain, we find that
character-aware models provide large gains on a novel spelling task
(WikiSpell). Applying our learnings to the visual domain, we train a suite of
image generation models, and show that character-aware variants outperform
their character-blind counterparts across a range of novel text rendering tasks
(our DrawText benchmark). Our models set a much higher state-of-the-art on
visual spelling, with 30+ point accuracy gains over competitors on rare words,
despite training on far fewer examples.
| no_new_dataset | 0.711437 |
2212.12110 | Wuqi Zhang | Wuqi Zhang, Lili Wei, Shing-Chi Cheung, Yepang Liu, Shuqing Li, Lu
Liu, Michael R. Lyu | Combatting Front-Running in Smart Contracts: Attack Mining, Benchmark
Construction and Vulnerability Detector Evaluation | Accepted as Journal First paper on the Transactions on Software
Engineering (TSE-2022-12-0547) | null | 10.1109/TSE.2023.3270117 | null | cs.SE | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Front-running attacks have been a major concern on the blockchain. Attackers
launch front-running attacks by inserting additional transactions before
upcoming victim transactions to manipulate victim transaction executions and
make profits. Recent studies have shown that front-running attacks are
prevalent on the Ethereum blockchain and have caused millions of US dollars
loss. Vulnerable smart contracts, blockchain programs invoked by transactions,
are held responsible for front-running attacks. Although techniques to detect
front-running vulnerabilities have been proposed, their performance on
real-world vulnerable contracts is unclear. There is no large-scale benchmark
based on real attacks to evaluate their capabilities. This motivates us to
build a benchmark consisting of 513 real-world attacks with vulnerable code
labeled in 235 distinct smart contracts. We propose automated techniques to
effectively collect real-world attacks and localize the corresponding
vulnerable code at scale. Our experiments show that our approaches are
effective, achieving higher recall in finding real attacks and higher precision
in pinpointing vulnerabilities compared to the existing techniques. The
evaluation of seven state-of-the-art vulnerability detection techniques on the
benchmark reveals their inadequacy in detecting front-running vulnerabilities,
with a low recall of at most 6.04%. Our further analysis identifies four common
limitations in existing techniques: lack of support for inter-contract
analysis, inefficient constraint solving for cryptographic operations, improper
vulnerability patterns, and lack of token support.
| [
{
"version": "v1",
"created": "Fri, 23 Dec 2022 02:26:46 GMT"
},
{
"version": "v2",
"created": "Thu, 29 Dec 2022 13:04:54 GMT"
},
{
"version": "v3",
"created": "Wed, 19 Apr 2023 21:33:20 GMT"
},
{
"version": "v4",
"created": "Tue, 2 May 2023 23:11:33 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Zhang",
"Wuqi",
""
],
[
"Wei",
"Lili",
""
],
[
"Cheung",
"Shing-Chi",
""
],
[
"Liu",
"Yepang",
""
],
[
"Li",
"Shuqing",
""
],
[
"Liu",
"Lu",
""
],
[
"Lyu",
"Michael R.",
""
]
] | TITLE: Combatting Front-Running in Smart Contracts: Attack Mining, Benchmark
Construction and Vulnerability Detector Evaluation
ABSTRACT: Front-running attacks have been a major concern on the blockchain. Attackers
launch front-running attacks by inserting additional transactions before
upcoming victim transactions to manipulate victim transaction executions and
make profits. Recent studies have shown that front-running attacks are
prevalent on the Ethereum blockchain and have caused millions of US dollars
loss. Vulnerable smart contracts, blockchain programs invoked by transactions,
are held responsible for front-running attacks. Although techniques to detect
front-running vulnerabilities have been proposed, their performance on
real-world vulnerable contracts is unclear. There is no large-scale benchmark
based on real attacks to evaluate their capabilities. This motivates us to
build a benchmark consisting of 513 real-world attacks with vulnerable code
labeled in 235 distinct smart contracts. We propose automated techniques to
effectively collect real-world attacks and localize the corresponding
vulnerable code at scale. Our experiments show that our approaches are
effective, achieving higher recall in finding real attacks and higher precision
in pinpointing vulnerabilities compared to the existing techniques. The
evaluation of seven state-of-the-art vulnerability detection techniques on the
benchmark reveals their inadequacy in detecting front-running vulnerabilities,
with a low recall of at most 6.04%. Our further analysis identifies four common
limitations in existing techniques: lack of support for inter-contract
analysis, inefficient constraint solving for cryptographic operations, improper
vulnerability patterns, and lack of token support.
| new_dataset | 0.707998 |
2301.06454 | Jonas Ditz | Jonas Christian Ditz and Jacqueline Wistuba-Hamprecht and Timo Maier
and Rolf Fendel and Nico Pfeifer and Bernhard Reuter | PlasmoFAB: A Benchmark to Foster Machine Learning for Plasmodium
falciparum Protein Antigen Candidate Prediction | null | null | null | null | q-bio.QM cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | Motivation: Machine learning methods can be used to support scientific
discovery in healthcare-related research fields. However, these methods can
only be reliably used if they can be trained on high-quality and curated
datasets. Currently, no such dataset for the exploration of Plasmodium
falciparum protein antigen candidates exists. The parasite Plasmodium
falciparum causes the infectious disease malaria. Thus, identifying potential
antigens is of utmost importance for the development of antimalarial drugs and
vaccines. Since exploring antigen candidates experimentally is an expensive and
time-consuming process, applying machine learning methods to support this
process has the potential to accelerate the development of drugs and vaccines,
which are needed for fighting and controlling malaria. Results: We developed
PlasmoFAB, a curated benchmark that can be used to train machine learning
methods for the exploration of Plasmodium falciparum protein antigen
candidates. We combined an extensive literature search with domain expertise to
create high-quality labels for Plasmodium falciparum specific proteins that
distinguish between antigen candidates and intracellular proteins.
Additionally, we used our benchmark to compare different well-known prediction
models and available protein localization prediction services on the task of
identifying protein antigen candidates. We show that available general-purpose
services are unable to provide sufficient performance on identifying protein
antigen candidates and are outperformed by our models that were trained on this
tailored data. Availability: PlasmoFAB is publicly available on Zenodo with DOI
10.5281/zenodo.7433087. Furthermore, all scripts that were used in the creation
of PlasmoFAB and the training and evaluation of machine learning models are
open source and publicly available on GitHub here:
https://github.com/msmdev/PlasmoFAB.
| [
{
"version": "v1",
"created": "Mon, 16 Jan 2023 14:57:42 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 11:41:56 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Ditz",
"Jonas Christian",
""
],
[
"Wistuba-Hamprecht",
"Jacqueline",
""
],
[
"Maier",
"Timo",
""
],
[
"Fendel",
"Rolf",
""
],
[
"Pfeifer",
"Nico",
""
],
[
"Reuter",
"Bernhard",
""
]
] | TITLE: PlasmoFAB: A Benchmark to Foster Machine Learning for Plasmodium
falciparum Protein Antigen Candidate Prediction
ABSTRACT: Motivation: Machine learning methods can be used to support scientific
discovery in healthcare-related research fields. However, these methods can
only be reliably used if they can be trained on high-quality and curated
datasets. Currently, no such dataset for the exploration of Plasmodium
falciparum protein antigen candidates exists. The parasite Plasmodium
falciparum causes the infectious disease malaria. Thus, identifying potential
antigens is of utmost importance for the development of antimalarial drugs and
vaccines. Since exploring antigen candidates experimentally is an expensive and
time-consuming process, applying machine learning methods to support this
process has the potential to accelerate the development of drugs and vaccines,
which are needed for fighting and controlling malaria. Results: We developed
PlasmoFAB, a curated benchmark that can be used to train machine learning
methods for the exploration of Plasmodium falciparum protein antigen
candidates. We combined an extensive literature search with domain expertise to
create high-quality labels for Plasmodium falciparum specific proteins that
distinguish between antigen candidates and intracellular proteins.
Additionally, we used our benchmark to compare different well-known prediction
models and available protein localization prediction services on the task of
identifying protein antigen candidates. We show that available general-purpose
services are unable to provide sufficient performance on identifying protein
antigen candidates and are outperformed by our models that were trained on this
tailored data. Availability: PlasmoFAB is publicly available on Zenodo with DOI
10.5281/zenodo.7433087. Furthermore, all scripts that were used in the creation
of PlasmoFAB and the training and evaluation of machine learning models are
open source and publicly available on GitHub here:
https://github.com/msmdev/PlasmoFAB.
| new_dataset | 0.718773 |
2302.05658 | Milan \v{S}ulc | \v{S}t\v{e}p\'an \v{S}imsa and Milan \v{S}ulc and Michal
U\v{r}i\v{c}\'a\v{r} and Yash Patel and Ahmed Hamdi and Mat\v{e}j Koci\'an
and Maty\'a\v{s} Skalick\'y and Ji\v{r}\'i Matas and Antoine Doucet and
Micka\"el Coustaty and Dimosthenis Karatzas | DocILE Benchmark for Document Information Localization and Extraction | Accepted to ICDAR 2023 | null | null | null | cs.CL cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper introduces the DocILE benchmark with the largest dataset of
business documents for the tasks of Key Information Localization and Extraction
and Line Item Recognition. It contains 6.7k annotated business documents, 100k
synthetically generated documents, and nearly~1M unlabeled documents for
unsupervised pre-training. The dataset has been built with knowledge of domain-
and task-specific aspects, resulting in the following key features: (i)
annotations in 55 classes, which surpasses the granularity of previously
published key information extraction datasets by a large margin; (ii) Line Item
Recognition represents a highly practical information extraction task, where
key information has to be assigned to items in a table; (iii) documents come
from numerous layouts and the test set includes zero- and few-shot cases as
well as layouts commonly seen in the training set. The benchmark comes with
several baselines, including RoBERTa, LayoutLMv3 and DETR-based Table
Transformer; applied to both tasks of the DocILE benchmark, with results shared
in this paper, offering a quick starting point for future work. The dataset,
baselines and supplementary material are available at
https://github.com/rossumai/docile.
| [
{
"version": "v1",
"created": "Sat, 11 Feb 2023 11:32:10 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 16:24:58 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Šimsa",
"Štěpán",
""
],
[
"Šulc",
"Milan",
""
],
[
"Uřičář",
"Michal",
""
],
[
"Patel",
"Yash",
""
],
[
"Hamdi",
"Ahmed",
""
],
[
"Kocián",
"Matěj",
""
],
[
"Skalický",
"Matyáš",
""
],
[
"Matas",
"Jiří",
""
],
[
"Doucet",
"Antoine",
""
],
[
"Coustaty",
"Mickaël",
""
],
[
"Karatzas",
"Dimosthenis",
""
]
] | TITLE: DocILE Benchmark for Document Information Localization and Extraction
ABSTRACT: This paper introduces the DocILE benchmark with the largest dataset of
business documents for the tasks of Key Information Localization and Extraction
and Line Item Recognition. It contains 6.7k annotated business documents, 100k
synthetically generated documents, and nearly~1M unlabeled documents for
unsupervised pre-training. The dataset has been built with knowledge of domain-
and task-specific aspects, resulting in the following key features: (i)
annotations in 55 classes, which surpasses the granularity of previously
published key information extraction datasets by a large margin; (ii) Line Item
Recognition represents a highly practical information extraction task, where
key information has to be assigned to items in a table; (iii) documents come
from numerous layouts and the test set includes zero- and few-shot cases as
well as layouts commonly seen in the training set. The benchmark comes with
several baselines, including RoBERTa, LayoutLMv3 and DETR-based Table
Transformer; applied to both tasks of the DocILE benchmark, with results shared
in this paper, offering a quick starting point for future work. The dataset,
baselines and supplementary material are available at
https://github.com/rossumai/docile.
| new_dataset | 0.718798 |
2302.07344 | Levi Cai | Levi Cai and Nathan E. McGuire and Roger Hanlon and T. Aran Mooney and
Yogesh Girdhar | Semi-Supervised Visual Tracking of Marine Animals using Autonomous
Underwater Vehicles | To appear in IJCV SI: Animal Tracking | Cai, Levi, Nathan E. McGuire, Roger Hanlon, T. Aran Mooney, and
Yogesh Girdhar. "Semi-supervised Visual Tracking of Marine Animals Using
Autonomous Underwater Vehicles." International Journal of Computer Vision
(2023): 1-22 | 10.1007/s11263-023-01762-5 | null | cs.CV cs.LG cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In-situ visual observations of marine organisms is crucial to developing
behavioural understandings and their relations to their surrounding ecosystem.
Typically, these observations are collected via divers, tags, and
remotely-operated or human-piloted vehicles. Recently, however, autonomous
underwater vehicles equipped with cameras and embedded computers with GPU
capabilities are being developed for a variety of applications, and in
particular, can be used to supplement these existing data collection mechanisms
where human operation or tags are more difficult. Existing approaches have
focused on using fully-supervised tracking methods, but labelled data for many
underwater species are severely lacking. Semi-supervised trackers may offer
alternative tracking solutions because they require less data than
fully-supervised counterparts. However, because there are not existing
realistic underwater tracking datasets, the performance of semi-supervised
tracking algorithms in the marine domain is not well understood. To better
evaluate their performance and utility, in this paper we provide (1) a novel
dataset specific to marine animals located at http://warp.whoi.edu/vmat/, (2)
an evaluation of state-of-the-art semi-supervised algorithms in the context of
underwater animal tracking, and (3) an evaluation of real-world performance
through demonstrations using a semi-supervised algorithm on-board an autonomous
underwater vehicle to track marine animals in the wild.
| [
{
"version": "v1",
"created": "Tue, 14 Feb 2023 21:08:52 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Cai",
"Levi",
""
],
[
"McGuire",
"Nathan E.",
""
],
[
"Hanlon",
"Roger",
""
],
[
"Mooney",
"T. Aran",
""
],
[
"Girdhar",
"Yogesh",
""
]
] | TITLE: Semi-Supervised Visual Tracking of Marine Animals using Autonomous
Underwater Vehicles
ABSTRACT: In-situ visual observations of marine organisms is crucial to developing
behavioural understandings and their relations to their surrounding ecosystem.
Typically, these observations are collected via divers, tags, and
remotely-operated or human-piloted vehicles. Recently, however, autonomous
underwater vehicles equipped with cameras and embedded computers with GPU
capabilities are being developed for a variety of applications, and in
particular, can be used to supplement these existing data collection mechanisms
where human operation or tags are more difficult. Existing approaches have
focused on using fully-supervised tracking methods, but labelled data for many
underwater species are severely lacking. Semi-supervised trackers may offer
alternative tracking solutions because they require less data than
fully-supervised counterparts. However, because there are not existing
realistic underwater tracking datasets, the performance of semi-supervised
tracking algorithms in the marine domain is not well understood. To better
evaluate their performance and utility, in this paper we provide (1) a novel
dataset specific to marine animals located at http://warp.whoi.edu/vmat/, (2)
an evaluation of state-of-the-art semi-supervised algorithms in the context of
underwater animal tracking, and (3) an evaluation of real-world performance
through demonstrations using a semi-supervised algorithm on-board an autonomous
underwater vehicle to track marine animals in the wild.
| new_dataset | 0.720175 |
2302.08092 | Yang Deng | Yang Deng, Wenxuan Zhang, Qian Yu, Wai Lam | Product Question Answering in E-Commerce: A Survey | Accepted by ACL 2023 main conference | null | null | null | cs.CL cs.IR | http://creativecommons.org/licenses/by/4.0/ | Product question answering (PQA), aiming to automatically provide instant
responses to customer's questions in E-Commerce platforms, has drawn increasing
attention in recent years. Compared with typical QA problems, PQA exhibits
unique challenges such as the subjectivity and reliability of user-generated
contents in E-commerce platforms. Therefore, various problem settings and novel
methods have been proposed to capture these special characteristics. In this
paper, we aim to systematically review existing research efforts on PQA.
Specifically, we categorize PQA studies into four problem settings in terms of
the form of provided answers. We analyze the pros and cons, as well as present
existing datasets and evaluation protocols for each setting. We further
summarize the most significant challenges that characterize PQA from general QA
applications and discuss their corresponding solutions. Finally, we conclude
this paper by providing the prospect on several future directions.
| [
{
"version": "v1",
"created": "Thu, 16 Feb 2023 05:09:41 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 07:29:41 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Deng",
"Yang",
""
],
[
"Zhang",
"Wenxuan",
""
],
[
"Yu",
"Qian",
""
],
[
"Lam",
"Wai",
""
]
] | TITLE: Product Question Answering in E-Commerce: A Survey
ABSTRACT: Product question answering (PQA), aiming to automatically provide instant
responses to customer's questions in E-Commerce platforms, has drawn increasing
attention in recent years. Compared with typical QA problems, PQA exhibits
unique challenges such as the subjectivity and reliability of user-generated
contents in E-commerce platforms. Therefore, various problem settings and novel
methods have been proposed to capture these special characteristics. In this
paper, we aim to systematically review existing research efforts on PQA.
Specifically, we categorize PQA studies into four problem settings in terms of
the form of provided answers. We analyze the pros and cons, as well as present
existing datasets and evaluation protocols for each setting. We further
summarize the most significant challenges that characterize PQA from general QA
applications and discuss their corresponding solutions. Finally, we conclude
this paper by providing the prospect on several future directions.
| no_new_dataset | 0.707847 |
2302.08676 | Su Feng | Su Feng, Boris Glavic, Oliver Kennedy | Efficient Approximation of Certain and Possible Answers for Ranking and
Window Queries over Uncertain Data (Extended version) | null | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Uncertainty arises naturally inmany application domains due to, e.g., data
entry errors and ambiguity in data cleaning. Prior work in incomplete and
probabilistic databases has investigated the semantics and efficient evaluation
of ranking and top-k queries over uncertain data. However, most approaches deal
with top-k and ranking in isolation and do represent uncertain input data and
query results using separate, incompatible datamodels. We present an efficient
approach for under- and over-approximating results of ranking, top-k, and
window queries over uncertain data. Our approach integrates well with existing
techniques for querying uncertain data, is efficient, and is to the best of our
knowledge the first to support windowed aggregation. We design algorithms for
physical operators for uncertain sorting and windowed aggregation, and
implement them in PostgreSQL.We evaluated our approach on synthetic and real
world datasets, demonstrating that it outperforms all competitors, and often
produces more accurate results.
| [
{
"version": "v1",
"created": "Fri, 17 Feb 2023 03:56:56 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 07:59:20 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Feng",
"Su",
""
],
[
"Glavic",
"Boris",
""
],
[
"Kennedy",
"Oliver",
""
]
] | TITLE: Efficient Approximation of Certain and Possible Answers for Ranking and
Window Queries over Uncertain Data (Extended version)
ABSTRACT: Uncertainty arises naturally inmany application domains due to, e.g., data
entry errors and ambiguity in data cleaning. Prior work in incomplete and
probabilistic databases has investigated the semantics and efficient evaluation
of ranking and top-k queries over uncertain data. However, most approaches deal
with top-k and ranking in isolation and do represent uncertain input data and
query results using separate, incompatible datamodels. We present an efficient
approach for under- and over-approximating results of ranking, top-k, and
window queries over uncertain data. Our approach integrates well with existing
techniques for querying uncertain data, is efficient, and is to the best of our
knowledge the first to support windowed aggregation. We design algorithms for
physical operators for uncertain sorting and windowed aggregation, and
implement them in PostgreSQL.We evaluated our approach on synthetic and real
world datasets, demonstrating that it outperforms all competitors, and often
produces more accurate results.
| no_new_dataset | 0.709025 |
2302.09173 | Lajanugen Logeswaran | Lajanugen Logeswaran, Sungryull Sohn, Yunseok Jang, Moontae Lee,
Honglak Lee | Unsupervised Task Graph Generation from Instructional Video Transcripts | Findings of ACL 2023 | null | null | null | cs.AI cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work explores the problem of generating task graphs of real-world
activities. Different from prior formulations, we consider a setting where text
transcripts of instructional videos performing a real-world activity (e.g.,
making coffee) are provided and the goal is to identify the key steps relevant
to the task as well as the dependency relationship between these key steps. We
propose a novel task graph generation approach that combines the reasoning
capabilities of instruction-tuned language models along with clustering and
ranking components to generate accurate task graphs in a completely
unsupervised manner. We show that the proposed approach generates more accurate
task graphs compared to a supervised learning approach on tasks from the ProceL
and CrossTask datasets.
| [
{
"version": "v1",
"created": "Fri, 17 Feb 2023 22:50:08 GMT"
},
{
"version": "v2",
"created": "Tue, 2 May 2023 19:46:14 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Logeswaran",
"Lajanugen",
""
],
[
"Sohn",
"Sungryull",
""
],
[
"Jang",
"Yunseok",
""
],
[
"Lee",
"Moontae",
""
],
[
"Lee",
"Honglak",
""
]
] | TITLE: Unsupervised Task Graph Generation from Instructional Video Transcripts
ABSTRACT: This work explores the problem of generating task graphs of real-world
activities. Different from prior formulations, we consider a setting where text
transcripts of instructional videos performing a real-world activity (e.g.,
making coffee) are provided and the goal is to identify the key steps relevant
to the task as well as the dependency relationship between these key steps. We
propose a novel task graph generation approach that combines the reasoning
capabilities of instruction-tuned language models along with clustering and
ranking components to generate accurate task graphs in a completely
unsupervised manner. We show that the proposed approach generates more accurate
task graphs compared to a supervised learning approach on tasks from the ProceL
and CrossTask datasets.
| no_new_dataset | 0.711249 |
2302.11362 | Yuchen Hu | Yuchen Hu, Chen Chen, Ruizhe Li, Qiushi Zhu, Eng Siong Chng | Gradient Remedy for Multi-Task Learning in End-to-End Noise-Robust
Speech Recognition | 5 pages, 5 figures, Accepted by ICASSP 2023 | null | null | null | eess.AS cs.LG cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Speech enhancement (SE) is proved effective in reducing noise from noisy
speech signals for downstream automatic speech recognition (ASR), where
multi-task learning strategy is employed to jointly optimize these two tasks.
However, the enhanced speech learned by SE objective may not always yield good
ASR results. From the optimization view, there sometimes exists interference
between the gradients of SE and ASR tasks, which could hinder the multi-task
learning and finally lead to sub-optimal ASR performance. In this paper, we
propose a simple yet effective approach called gradient remedy (GR) to solve
interference between task gradients in noise-robust speech recognition, from
perspectives of both angle and magnitude. Specifically, we first project the SE
task's gradient onto a dynamic surface that is at acute angle to ASR gradient,
in order to remove the conflict between them and assist in ASR optimization.
Furthermore, we adaptively rescale the magnitude of two gradients to prevent
the dominant ASR task from being misled by SE gradient. Experimental results
show that the proposed approach well resolves the gradient interference and
achieves relative word error rate (WER) reductions of 9.3% and 11.1% over
multi-task learning baseline, on RATS and CHiME-4 datasets, respectively. Our
code is available at GitHub.
| [
{
"version": "v1",
"created": "Wed, 22 Feb 2023 13:31:13 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 05:06:51 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Hu",
"Yuchen",
""
],
[
"Chen",
"Chen",
""
],
[
"Li",
"Ruizhe",
""
],
[
"Zhu",
"Qiushi",
""
],
[
"Chng",
"Eng Siong",
""
]
] | TITLE: Gradient Remedy for Multi-Task Learning in End-to-End Noise-Robust
Speech Recognition
ABSTRACT: Speech enhancement (SE) is proved effective in reducing noise from noisy
speech signals for downstream automatic speech recognition (ASR), where
multi-task learning strategy is employed to jointly optimize these two tasks.
However, the enhanced speech learned by SE objective may not always yield good
ASR results. From the optimization view, there sometimes exists interference
between the gradients of SE and ASR tasks, which could hinder the multi-task
learning and finally lead to sub-optimal ASR performance. In this paper, we
propose a simple yet effective approach called gradient remedy (GR) to solve
interference between task gradients in noise-robust speech recognition, from
perspectives of both angle and magnitude. Specifically, we first project the SE
task's gradient onto a dynamic surface that is at acute angle to ASR gradient,
in order to remove the conflict between them and assist in ASR optimization.
Furthermore, we adaptively rescale the magnitude of two gradients to prevent
the dominant ASR task from being misled by SE gradient. Experimental results
show that the proposed approach well resolves the gradient interference and
achieves relative word error rate (WER) reductions of 9.3% and 11.1% over
multi-task learning baseline, on RATS and CHiME-4 datasets, respectively. Our
code is available at GitHub.
| no_new_dataset | 0.708244 |
2302.14320 | Monica Welfert | Monica Welfert, Kyle Otstot, Gowtham R. Kurri, Lalitha Sankar | $(\alpha_D,\alpha_G)$-GANs: Addressing GAN Training Instabilities via
Dual Objectives | Extended version of a paper accepted to ISIT 2023 | null | null | null | cs.LG cs.IT math.IT stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In an effort to address the training instabilities of GANs, we introduce a
class of dual-objective GANs with different value functions (objectives) for
the generator (G) and discriminator (D). In particular, we model each objective
using $\alpha$-loss, a tunable classification loss, to obtain
$(\alpha_D,\alpha_G)$-GANs, parameterized by $(\alpha_D,\alpha_G)\in
(0,\infty]^2$. For sufficiently large number of samples and capacities for G
and D, we show that the resulting non-zero sum game simplifies to minimizing an
$f$-divergence under appropriate conditions on $(\alpha_D,\alpha_G)$. In the
finite sample and capacity setting, we define estimation error to quantify the
gap in the generator's performance relative to the optimal setting with
infinite samples and obtain upper bounds on this error, showing it to be order
optimal under certain conditions. Finally, we highlight the value of tuning
$(\alpha_D,\alpha_G)$ in alleviating training instabilities for the synthetic
2D Gaussian mixture ring and the Stacked MNIST datasets.
| [
{
"version": "v1",
"created": "Tue, 28 Feb 2023 05:22:54 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 04:22:14 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Welfert",
"Monica",
""
],
[
"Otstot",
"Kyle",
""
],
[
"Kurri",
"Gowtham R.",
""
],
[
"Sankar",
"Lalitha",
""
]
] | TITLE: $(\alpha_D,\alpha_G)$-GANs: Addressing GAN Training Instabilities via
Dual Objectives
ABSTRACT: In an effort to address the training instabilities of GANs, we introduce a
class of dual-objective GANs with different value functions (objectives) for
the generator (G) and discriminator (D). In particular, we model each objective
using $\alpha$-loss, a tunable classification loss, to obtain
$(\alpha_D,\alpha_G)$-GANs, parameterized by $(\alpha_D,\alpha_G)\in
(0,\infty]^2$. For sufficiently large number of samples and capacities for G
and D, we show that the resulting non-zero sum game simplifies to minimizing an
$f$-divergence under appropriate conditions on $(\alpha_D,\alpha_G)$. In the
finite sample and capacity setting, we define estimation error to quantify the
gap in the generator's performance relative to the optimal setting with
infinite samples and obtain upper bounds on this error, showing it to be order
optimal under certain conditions. Finally, we highlight the value of tuning
$(\alpha_D,\alpha_G)$ in alleviating training instabilities for the synthetic
2D Gaussian mixture ring and the Stacked MNIST datasets.
| no_new_dataset | 0.707209 |
2303.07519 | Antonios Liapis | Theodoros Galanos, Antonios Liapis and Georgios N. Yannakakis | Architext: Language-Driven Generative Architecture Design | 21 pages | null | null | null | cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Architectural design is a highly complex practice that involves a wide
diversity of disciplines, technologies, proprietary design software, expertise,
and an almost infinite number of constraints, across a vast array of design
tasks. Enabling intuitive, accessible, and scalable design processes is an
important step towards performance-driven and sustainable design for all. To
that end, we introduce Architext, a novel semantic generation assistive tool.
Architext enables design generation with only natural language prompts, given
to large-scale Language Models, as input. We conduct a thorough quantitative
evaluation of Architext's downstream task performance, focusing on semantic
accuracy and diversity for a number of pre-trained language models ranging from
120 million to 6 billion parameters. Architext models are able to learn the
specific design task, generating valid residential layouts at a near 100% rate.
Accuracy shows great improvement when scaling the models, with the largest
model (GPT-J) yielding impressive accuracy ranging between 25% to over 80% for
different prompt categories. We open source the finetuned Architext models and
our synthetic dataset, hoping to inspire experimentation in this exciting area
of design research.
| [
{
"version": "v1",
"created": "Mon, 13 Mar 2023 23:11:05 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 16:07:05 GMT"
},
{
"version": "v3",
"created": "Wed, 3 May 2023 09:29:05 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Galanos",
"Theodoros",
""
],
[
"Liapis",
"Antonios",
""
],
[
"Yannakakis",
"Georgios N.",
""
]
] | TITLE: Architext: Language-Driven Generative Architecture Design
ABSTRACT: Architectural design is a highly complex practice that involves a wide
diversity of disciplines, technologies, proprietary design software, expertise,
and an almost infinite number of constraints, across a vast array of design
tasks. Enabling intuitive, accessible, and scalable design processes is an
important step towards performance-driven and sustainable design for all. To
that end, we introduce Architext, a novel semantic generation assistive tool.
Architext enables design generation with only natural language prompts, given
to large-scale Language Models, as input. We conduct a thorough quantitative
evaluation of Architext's downstream task performance, focusing on semantic
accuracy and diversity for a number of pre-trained language models ranging from
120 million to 6 billion parameters. Architext models are able to learn the
specific design task, generating valid residential layouts at a near 100% rate.
Accuracy shows great improvement when scaling the models, with the largest
model (GPT-J) yielding impressive accuracy ranging between 25% to over 80% for
different prompt categories. We open source the finetuned Architext models and
our synthetic dataset, hoping to inspire experimentation in this exciting area
of design research.
| new_dataset | 0.712201 |
2304.07302 | Qijie Bai | Qijie Bai, Changli Nie, Haiwei Zhang, Dongming Zhao, Xiaojie Yuan | HGWaveNet: A Hyperbolic Graph Neural Network for Temporal Link
Prediction | Accepted by Web Conference (WWW) 2023 | WWW '23: Proceedings of the ACM Web Conference 2023 (523-532) | 10.1145/3543507.3583455 | null | cs.LG cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Temporal link prediction, aiming to predict future edges between paired nodes
in a dynamic graph, is of vital importance in diverse applications. However,
existing methods are mainly built upon uniform Euclidean space, which has been
found to be conflict with the power-law distributions of real-world graphs and
unable to represent the hierarchical connections between nodes effectively.
With respect to the special data characteristic, hyperbolic geometry offers an
ideal alternative due to its exponential expansion property. In this paper, we
propose HGWaveNet, a novel hyperbolic graph neural network that fully exploits
the fitness between hyperbolic spaces and data distributions for temporal link
prediction. Specifically, we design two key modules to learn the spatial
topological structures and temporal evolutionary information separately. On the
one hand, a hyperbolic diffusion graph convolution (HDGC) module effectively
aggregates information from a wider range of neighbors. On the other hand, the
internal order of causal correlation between historical states is captured by
hyperbolic dilated causal convolution (HDCC) modules. The whole model is built
upon the hyperbolic spaces to preserve the hierarchical structural information
in the entire data flow. To prove the superiority of HGWaveNet, extensive
experiments are conducted on six real-world graph datasets and the results show
a relative improvement by up to 6.67% on AUC for temporal link prediction over
SOTA methods.
| [
{
"version": "v1",
"created": "Fri, 14 Apr 2023 07:07:00 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 04:54:52 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Bai",
"Qijie",
""
],
[
"Nie",
"Changli",
""
],
[
"Zhang",
"Haiwei",
""
],
[
"Zhao",
"Dongming",
""
],
[
"Yuan",
"Xiaojie",
""
]
] | TITLE: HGWaveNet: A Hyperbolic Graph Neural Network for Temporal Link
Prediction
ABSTRACT: Temporal link prediction, aiming to predict future edges between paired nodes
in a dynamic graph, is of vital importance in diverse applications. However,
existing methods are mainly built upon uniform Euclidean space, which has been
found to be conflict with the power-law distributions of real-world graphs and
unable to represent the hierarchical connections between nodes effectively.
With respect to the special data characteristic, hyperbolic geometry offers an
ideal alternative due to its exponential expansion property. In this paper, we
propose HGWaveNet, a novel hyperbolic graph neural network that fully exploits
the fitness between hyperbolic spaces and data distributions for temporal link
prediction. Specifically, we design two key modules to learn the spatial
topological structures and temporal evolutionary information separately. On the
one hand, a hyperbolic diffusion graph convolution (HDGC) module effectively
aggregates information from a wider range of neighbors. On the other hand, the
internal order of causal correlation between historical states is captured by
hyperbolic dilated causal convolution (HDCC) modules. The whole model is built
upon the hyperbolic spaces to preserve the hierarchical structural information
in the entire data flow. To prove the superiority of HGWaveNet, extensive
experiments are conducted on six real-world graph datasets and the results show
a relative improvement by up to 6.67% on AUC for temporal link prediction over
SOTA methods.
| no_new_dataset | 0.709435 |
2304.07334 | Dingwen Tao | Chengming Zhang, Shaden Smith, Baixi Sun, Jiannan Tian, Jonathan
Soifer, Xiaodong Yu, Shuaiwen Leon Song, Yuxiong He, Dingwen Tao | HEAT: A Highly Efficient and Affordable Training System for
Collaborative Filtering Based Recommendation on CPUs | 12 pages, 14 figures, 7 tables, accepted by ACM ICS '23 | null | 10.1145/3577193.3593717 | null | cs.DC cs.LG | http://creativecommons.org/licenses/by/4.0/ | Collaborative filtering (CF) has been proven to be one of the most effective
techniques for recommendation. Among all CF approaches, SimpleX is the
state-of-the-art method that adopts a novel loss function and a proper number
of negative samples. However, there is no work that optimizes SimpleX on
multi-core CPUs, leading to limited performance. To this end, we perform an
in-depth profiling and analysis of existing SimpleX implementations and
identify their performance bottlenecks including (1) irregular memory accesses,
(2) unnecessary memory copies, and (3) redundant computations. To address these
issues, we propose an efficient CF training system (called HEAT) that fully
enables the multi-level caching and multi-threading capabilities of modern
CPUs. Specifically, the optimization of HEAT is threefold: (1) It tiles the
embedding matrix to increase data locality and reduce cache misses (thus
reduces read latency); (2) It optimizes stochastic gradient descent (SGD) with
sampling by parallelizing vector products instead of matrix-matrix
multiplications, in particular the similarity computation therein, to avoid
memory copies for matrix data preparation; and (3) It aggressively reuses
intermediate results from the forward phase in the backward phase to alleviate
redundant computation. Evaluation on five widely used datasets with both x86-
and ARM-architecture processors shows that HEAT achieves up to 45.2X speedup
over existing CPU solution and 4.5X speedup and 7.9X cost reduction in Cloud
over existing GPU solution with NVIDIA V100 GPU.
| [
{
"version": "v1",
"created": "Fri, 14 Apr 2023 18:07:26 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 15:42:12 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Zhang",
"Chengming",
""
],
[
"Smith",
"Shaden",
""
],
[
"Sun",
"Baixi",
""
],
[
"Tian",
"Jiannan",
""
],
[
"Soifer",
"Jonathan",
""
],
[
"Yu",
"Xiaodong",
""
],
[
"Song",
"Shuaiwen Leon",
""
],
[
"He",
"Yuxiong",
""
],
[
"Tao",
"Dingwen",
""
]
] | TITLE: HEAT: A Highly Efficient and Affordable Training System for
Collaborative Filtering Based Recommendation on CPUs
ABSTRACT: Collaborative filtering (CF) has been proven to be one of the most effective
techniques for recommendation. Among all CF approaches, SimpleX is the
state-of-the-art method that adopts a novel loss function and a proper number
of negative samples. However, there is no work that optimizes SimpleX on
multi-core CPUs, leading to limited performance. To this end, we perform an
in-depth profiling and analysis of existing SimpleX implementations and
identify their performance bottlenecks including (1) irregular memory accesses,
(2) unnecessary memory copies, and (3) redundant computations. To address these
issues, we propose an efficient CF training system (called HEAT) that fully
enables the multi-level caching and multi-threading capabilities of modern
CPUs. Specifically, the optimization of HEAT is threefold: (1) It tiles the
embedding matrix to increase data locality and reduce cache misses (thus
reduces read latency); (2) It optimizes stochastic gradient descent (SGD) with
sampling by parallelizing vector products instead of matrix-matrix
multiplications, in particular the similarity computation therein, to avoid
memory copies for matrix data preparation; and (3) It aggressively reuses
intermediate results from the forward phase in the backward phase to alleviate
redundant computation. Evaluation on five widely used datasets with both x86-
and ARM-architecture processors shows that HEAT achieves up to 45.2X speedup
over existing CPU solution and 4.5X speedup and 7.9X cost reduction in Cloud
over existing GPU solution with NVIDIA V100 GPU.
| no_new_dataset | 0.708591 |
2304.07342 | Dingwen Tao | Boyuan Zhang, Jiannan Tian, Sheng Di, Xiaodong Yu, Martin Swany,
Dingwen Tao, Franck Cappello | GPULZ: Optimizing LZSS Lossless Compression for Multi-byte Data on
Modern GPUs | 12 pages, 9 figures, 3 tables, accepted by ACM ICS '23 | null | 10.1145/3577193.3593706 | null | cs.DC | http://creativecommons.org/licenses/by/4.0/ | Today's graphics processing unit (GPU) applications produce vast volumes of
data, which are challenging to store and transfer efficiently. Thus, data
compression is becoming a critical technique to mitigate the storage burden and
communication cost. LZSS is the core algorithm in many widely used compressors,
such as Deflate. However, existing GPU-based LZSS compressors suffer from low
throughput due to the sequential nature of the LZSS algorithm. Moreover, many
GPU applications produce multi-byte data (e.g., int16/int32 index,
floating-point numbers), while the current LZSS compression only takes
single-byte data as input. To this end, in this work, we propose GPULZ, a
highly efficient LZSS compression on modern GPUs for multi-byte data. The
contribution of our work is fourfold: First, we perform an in-depth analysis of
existing LZ compressors for GPUs and investigate their main issues. Then, we
propose two main algorithm-level optimizations. Specifically, we (1) change
prefix sum from one pass to two passes and fuse multiple kernels to reduce data
movement between shared memory and global memory, and (2) optimize existing
pattern-matching approach for multi-byte symbols to reduce computation
complexity and explore longer repeated patterns. Third, we perform
architectural performance optimizations, such as maximizing shared memory
utilization by adapting data partitions to different GPU architectures.
Finally, we evaluate GPULZ on six datasets of various types with NVIDIA A100
and A4000 GPUs. Results show that GPULZ achieves up to 272.1X speedup on A4000
and up to 1.4X higher compression ratio compared to state-of-the-art solutions.
| [
{
"version": "v1",
"created": "Fri, 14 Apr 2023 18:23:25 GMT"
},
{
"version": "v2",
"created": "Tue, 2 May 2023 19:12:08 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Zhang",
"Boyuan",
""
],
[
"Tian",
"Jiannan",
""
],
[
"Di",
"Sheng",
""
],
[
"Yu",
"Xiaodong",
""
],
[
"Swany",
"Martin",
""
],
[
"Tao",
"Dingwen",
""
],
[
"Cappello",
"Franck",
""
]
] | TITLE: GPULZ: Optimizing LZSS Lossless Compression for Multi-byte Data on
Modern GPUs
ABSTRACT: Today's graphics processing unit (GPU) applications produce vast volumes of
data, which are challenging to store and transfer efficiently. Thus, data
compression is becoming a critical technique to mitigate the storage burden and
communication cost. LZSS is the core algorithm in many widely used compressors,
such as Deflate. However, existing GPU-based LZSS compressors suffer from low
throughput due to the sequential nature of the LZSS algorithm. Moreover, many
GPU applications produce multi-byte data (e.g., int16/int32 index,
floating-point numbers), while the current LZSS compression only takes
single-byte data as input. To this end, in this work, we propose GPULZ, a
highly efficient LZSS compression on modern GPUs for multi-byte data. The
contribution of our work is fourfold: First, we perform an in-depth analysis of
existing LZ compressors for GPUs and investigate their main issues. Then, we
propose two main algorithm-level optimizations. Specifically, we (1) change
prefix sum from one pass to two passes and fuse multiple kernels to reduce data
movement between shared memory and global memory, and (2) optimize existing
pattern-matching approach for multi-byte symbols to reduce computation
complexity and explore longer repeated patterns. Third, we perform
architectural performance optimizations, such as maximizing shared memory
utilization by adapting data partitions to different GPU architectures.
Finally, we evaluate GPULZ on six datasets of various types with NVIDIA A100
and A4000 GPUs. Results show that GPULZ achieves up to 272.1X speedup on A4000
and up to 1.4X higher compression ratio compared to state-of-the-art solutions.
| no_new_dataset | 0.706804 |
2304.08253 | Sanne Vrijenhoek | Sanne Vrijenhoek | Do you MIND? Reflections on the MIND dataset for research on diversity
in news recommendations | null | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | The MIND dataset is at the moment of writing the most extensive dataset
available for the research and development of news recommender systems. This
work analyzes the suitability of the dataset for research on diverse news
recommendations. On the one hand we analyze the effect the different steps in
the recommendation pipeline have on the distribution of article categories, and
on the other hand we check whether the supplied data would be sufficient for
more sophisticated diversity analysis. We conclude that while MIND is a great
step forward, there is still a lot of room for improvement.
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 13:14:24 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 14:13:52 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Vrijenhoek",
"Sanne",
""
]
] | TITLE: Do you MIND? Reflections on the MIND dataset for research on diversity
in news recommendations
ABSTRACT: The MIND dataset is at the moment of writing the most extensive dataset
available for the research and development of news recommender systems. This
work analyzes the suitability of the dataset for research on diverse news
recommendations. On the one hand we analyze the effect the different steps in
the recommendation pipeline have on the distribution of article categories, and
on the other hand we check whether the supplied data would be sufficient for
more sophisticated diversity analysis. We conclude that while MIND is a great
step forward, there is still a lot of room for improvement.
| no_new_dataset | 0.702568 |
2304.12557 | Dingwen Tao | Boyuan Zhang, Jiannan Tian, Sheng Di, Xiaodong Yu, Yunhe Feng, Xin
Liang, Dingwen Tao, Franck Cappello | FZ-GPU: A Fast and High-Ratio Lossy Compressor for Scientific Computing
Applications on GPUs | 14 pages, 12 figures, accepted by ACM HPDC '23 | null | 10.1145/3588195.3592994 | null | cs.DC | http://creativecommons.org/licenses/by/4.0/ | Today's large-scale scientific applications running on high-performance
computing (HPC) systems generate vast data volumes. Thus, data compression is
becoming a critical technique to mitigate the storage burden and data-movement
cost. However, existing lossy compressors for scientific data cannot achieve a
high compression ratio and throughput simultaneously, hindering their adoption
in many applications requiring fast compression, such as in-memory compression.
To this end, in this work, we develop a fast and high-ratio error-bounded lossy
compressor on GPUs for scientific data (called FZ-GPU). Specifically, we first
design a new compression pipeline that consists of fully parallelized
quantization, bitshuffle, and our newly designed fast encoding. Then, we
propose a series of deep architectural optimizations for each kernel in the
pipeline to take full advantage of CUDA architectures. We propose a warp-level
optimization to avoid data conflicts for bit-wise operations in bitshuffle,
maximize shared memory utilization, and eliminate unnecessary data movements by
fusing different compression kernels. Finally, we evaluate FZ-GPU on two NVIDIA
GPUs (i.e., A100 and RTX A4000) using six representative scientific datasets
from SDRBench. Results on the A100 GPU show that FZ-GPU achieves an average
speedup of 4.2X over cuSZ and an average speedup of 37.0X over a multi-threaded
CPU implementation of our algorithm under the same error bound. FZ-GPU also
achieves an average speedup of 2.3X and an average compression ratio
improvement of 2.0X over cuZFP under the same data distortion.
| [
{
"version": "v1",
"created": "Tue, 25 Apr 2023 03:55:25 GMT"
},
{
"version": "v2",
"created": "Tue, 2 May 2023 19:04:59 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Zhang",
"Boyuan",
""
],
[
"Tian",
"Jiannan",
""
],
[
"Di",
"Sheng",
""
],
[
"Yu",
"Xiaodong",
""
],
[
"Feng",
"Yunhe",
""
],
[
"Liang",
"Xin",
""
],
[
"Tao",
"Dingwen",
""
],
[
"Cappello",
"Franck",
""
]
] | TITLE: FZ-GPU: A Fast and High-Ratio Lossy Compressor for Scientific Computing
Applications on GPUs
ABSTRACT: Today's large-scale scientific applications running on high-performance
computing (HPC) systems generate vast data volumes. Thus, data compression is
becoming a critical technique to mitigate the storage burden and data-movement
cost. However, existing lossy compressors for scientific data cannot achieve a
high compression ratio and throughput simultaneously, hindering their adoption
in many applications requiring fast compression, such as in-memory compression.
To this end, in this work, we develop a fast and high-ratio error-bounded lossy
compressor on GPUs for scientific data (called FZ-GPU). Specifically, we first
design a new compression pipeline that consists of fully parallelized
quantization, bitshuffle, and our newly designed fast encoding. Then, we
propose a series of deep architectural optimizations for each kernel in the
pipeline to take full advantage of CUDA architectures. We propose a warp-level
optimization to avoid data conflicts for bit-wise operations in bitshuffle,
maximize shared memory utilization, and eliminate unnecessary data movements by
fusing different compression kernels. Finally, we evaluate FZ-GPU on two NVIDIA
GPUs (i.e., A100 and RTX A4000) using six representative scientific datasets
from SDRBench. Results on the A100 GPU show that FZ-GPU achieves an average
speedup of 4.2X over cuSZ and an average speedup of 37.0X over a multi-threaded
CPU implementation of our algorithm under the same error bound. FZ-GPU also
achieves an average speedup of 2.3X and an average compression ratio
improvement of 2.0X over cuZFP under the same data distortion.
| no_new_dataset | 0.708055 |
2304.12730 | Avishek Lahiri | Avishek Lahiri, Debarshi Kumar Sanyal, Imon Mukherjee | CitePrompt: Using Prompts to Identify Citation Intent in Scientific
Papers | Selected for publication at ACM/IEEE JOINT CONFERENCE ON DIGITAL
LIBRARIES 2023 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Citations in scientific papers not only help us trace the intellectual
lineage but also are a useful indicator of the scientific significance of the
work. Citation intents prove beneficial as they specify the role of the
citation in a given context. In this paper, we present CitePrompt, a framework
which uses the hitherto unexplored approach of prompt-based learning for
citation intent classification. We argue that with the proper choice of the
pretrained language model, the prompt template, and the prompt verbalizer, we
can not only get results that are better than or comparable to those obtained
with the state-of-the-art methods but also do it with much less exterior
information about the scientific document. We report state-of-the-art results
on the ACL-ARC dataset, and also show significant improvement on the SciCite
dataset over all baseline models except one. As suitably large labelled
datasets for citation intent classification can be quite hard to find, in a
first, we propose the conversion of this task to the few-shot and zero-shot
settings. For the ACL-ARC dataset, we report a 53.86% F1 score for the
zero-shot setting, which improves to 63.61% and 66.99% for the 5-shot and
10-shot settings, respectively.
| [
{
"version": "v1",
"created": "Tue, 25 Apr 2023 11:19:52 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 11:47:55 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Lahiri",
"Avishek",
""
],
[
"Sanyal",
"Debarshi Kumar",
""
],
[
"Mukherjee",
"Imon",
""
]
] | TITLE: CitePrompt: Using Prompts to Identify Citation Intent in Scientific
Papers
ABSTRACT: Citations in scientific papers not only help us trace the intellectual
lineage but also are a useful indicator of the scientific significance of the
work. Citation intents prove beneficial as they specify the role of the
citation in a given context. In this paper, we present CitePrompt, a framework
which uses the hitherto unexplored approach of prompt-based learning for
citation intent classification. We argue that with the proper choice of the
pretrained language model, the prompt template, and the prompt verbalizer, we
can not only get results that are better than or comparable to those obtained
with the state-of-the-art methods but also do it with much less exterior
information about the scientific document. We report state-of-the-art results
on the ACL-ARC dataset, and also show significant improvement on the SciCite
dataset over all baseline models except one. As suitably large labelled
datasets for citation intent classification can be quite hard to find, in a
first, we propose the conversion of this task to the few-shot and zero-shot
settings. For the ACL-ARC dataset, we report a 53.86% F1 score for the
zero-shot setting, which improves to 63.61% and 66.99% for the 5-shot and
10-shot settings, respectively.
| no_new_dataset | 0.711017 |
2304.13349 | Zenan Shi | Zenan Shi, Haipeng Chen, Long Chen and Dong Zhang | Discrepancy-Guided Reconstruction Learning for Image Forgery Detection | 12 pages, 8 figures | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a novel image forgery detection paradigm for
boosting the model learning capacity on both forgery-sensitive and genuine
compact visual patterns. Compared to the existing methods that only focus on
the discrepant-specific patterns (\eg, noises, textures, and frequencies), our
method has a greater generalization. Specifically, we first propose a
Discrepancy-Guided Encoder (DisGE) to extract forgery-sensitive visual
patterns. DisGE consists of two branches, where the mainstream backbone branch
is used to extract general semantic features, and the accessorial discrepant
external attention branch is used to extract explicit forgery cues. Besides, a
Double-Head Reconstruction (DouHR) module is proposed to enhance genuine
compact visual patterns in different granular spaces. Under DouHR, we further
introduce a Discrepancy-Aggregation Detector (DisAD) to aggregate these genuine
compact visual patterns, such that the forgery detection capability on unknown
patterns can be improved. Extensive experimental results on four challenging
datasets validate the effectiveness of our proposed method against
state-of-the-art competitors.
| [
{
"version": "v1",
"created": "Wed, 26 Apr 2023 07:40:43 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 12:50:10 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Shi",
"Zenan",
""
],
[
"Chen",
"Haipeng",
""
],
[
"Chen",
"Long",
""
],
[
"Zhang",
"Dong",
""
]
] | TITLE: Discrepancy-Guided Reconstruction Learning for Image Forgery Detection
ABSTRACT: In this paper, we propose a novel image forgery detection paradigm for
boosting the model learning capacity on both forgery-sensitive and genuine
compact visual patterns. Compared to the existing methods that only focus on
the discrepant-specific patterns (\eg, noises, textures, and frequencies), our
method has a greater generalization. Specifically, we first propose a
Discrepancy-Guided Encoder (DisGE) to extract forgery-sensitive visual
patterns. DisGE consists of two branches, where the mainstream backbone branch
is used to extract general semantic features, and the accessorial discrepant
external attention branch is used to extract explicit forgery cues. Besides, a
Double-Head Reconstruction (DouHR) module is proposed to enhance genuine
compact visual patterns in different granular spaces. Under DouHR, we further
introduce a Discrepancy-Aggregation Detector (DisAD) to aggregate these genuine
compact visual patterns, such that the forgery detection capability on unknown
patterns can be improved. Extensive experimental results on four challenging
datasets validate the effectiveness of our proposed method against
state-of-the-art competitors.
| no_new_dataset | 0.709629 |
2304.14125 | Sami Arja | Sami Arja, Alexandre Marcireau, Richard L. Balthazor, Matthew G.
McHarg, Saeed Afshar and Gregory Cohen | Density Invariant Contrast Maximization for Neuromorphic Earth
Observations | Accepted to 2023 CVPRW Workshop on Event-Based Vision | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Contrast maximization (CMax) techniques are widely used in event-based vision
systems to estimate the motion parameters of the camera and generate
high-contrast images. However, these techniques are noise-intolerance and
suffer from the multiple extrema problem which arises when the scene contains
more noisy events than structure, causing the contrast to be higher at multiple
locations. This makes the task of estimating the camera motion extremely
challenging, which is a problem for neuromorphic earth observation, because,
without a proper estimation of the motion parameters, it is not possible to
generate a map with high contrast, causing important details to be lost.
Similar methods that use CMax addressed this problem by changing or augmenting
the objective function to enable it to converge to the correct motion
parameters. Our proposed solution overcomes the multiple extrema and
noise-intolerance problems by correcting the warped event before calculating
the contrast and offers the following advantages: it does not depend on the
event data, it does not require a prior about the camera motion, and keeps the
rest of the CMax pipeline unchanged. This is to ensure that the contrast is
only high around the correct motion parameters. Our approach enables the
creation of better motion-compensated maps through an analytical compensation
technique using a novel dataset from the International Space Station (ISS).
Code is available at \url{https://github.com/neuromorphicsystems/event_warping}
| [
{
"version": "v1",
"created": "Thu, 27 Apr 2023 12:17:40 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 12:33:40 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Arja",
"Sami",
""
],
[
"Marcireau",
"Alexandre",
""
],
[
"Balthazor",
"Richard L.",
""
],
[
"McHarg",
"Matthew G.",
""
],
[
"Afshar",
"Saeed",
""
],
[
"Cohen",
"Gregory",
""
]
] | TITLE: Density Invariant Contrast Maximization for Neuromorphic Earth
Observations
ABSTRACT: Contrast maximization (CMax) techniques are widely used in event-based vision
systems to estimate the motion parameters of the camera and generate
high-contrast images. However, these techniques are noise-intolerance and
suffer from the multiple extrema problem which arises when the scene contains
more noisy events than structure, causing the contrast to be higher at multiple
locations. This makes the task of estimating the camera motion extremely
challenging, which is a problem for neuromorphic earth observation, because,
without a proper estimation of the motion parameters, it is not possible to
generate a map with high contrast, causing important details to be lost.
Similar methods that use CMax addressed this problem by changing or augmenting
the objective function to enable it to converge to the correct motion
parameters. Our proposed solution overcomes the multiple extrema and
noise-intolerance problems by correcting the warped event before calculating
the contrast and offers the following advantages: it does not depend on the
event data, it does not require a prior about the camera motion, and keeps the
rest of the CMax pipeline unchanged. This is to ensure that the contrast is
only high around the correct motion parameters. Our approach enables the
creation of better motion-compensated maps through an analytical compensation
technique using a novel dataset from the International Space Station (ISS).
Code is available at \url{https://github.com/neuromorphicsystems/event_warping}
| no_new_dataset | 0.648466 |
2305.01507 | Naoki Masuyama | Naoki Masuyama, Takanori Takebayashi, Yusuke Nojima, Chu Kiong Loo,
Hisao Ishibuchi, Stefan Wermter | A Parameter-free Adaptive Resonance Theory-based Topological Clustering
Algorithm Capable of Continual Learning | This paper is currently under review | null | null | null | cs.NE cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In general, a similarity threshold (i.e., a vigilance parameter) for a node
learning process in Adaptive Resonance Theory (ART)-based algorithms has a
significant impact on clustering performance. In addition, an edge deletion
threshold in a topological clustering algorithm plays an important role in
adaptively generating well-separated clusters during a self-organizing process.
In this paper, we propose a new parameter-free ART-based topological clustering
algorithm capable of continual learning by introducing parameter estimation
methods. Experimental results with synthetic and real-world datasets show that
the proposed algorithm has superior clustering performance to the
state-of-the-art clustering algorithms without any parameter
pre-specifications.
| [
{
"version": "v1",
"created": "Mon, 1 May 2023 01:04:07 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 01:58:25 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Masuyama",
"Naoki",
""
],
[
"Takebayashi",
"Takanori",
""
],
[
"Nojima",
"Yusuke",
""
],
[
"Loo",
"Chu Kiong",
""
],
[
"Ishibuchi",
"Hisao",
""
],
[
"Wermter",
"Stefan",
""
]
] | TITLE: A Parameter-free Adaptive Resonance Theory-based Topological Clustering
Algorithm Capable of Continual Learning
ABSTRACT: In general, a similarity threshold (i.e., a vigilance parameter) for a node
learning process in Adaptive Resonance Theory (ART)-based algorithms has a
significant impact on clustering performance. In addition, an edge deletion
threshold in a topological clustering algorithm plays an important role in
adaptively generating well-separated clusters during a self-organizing process.
In this paper, we propose a new parameter-free ART-based topological clustering
algorithm capable of continual learning by introducing parameter estimation
methods. Experimental results with synthetic and real-world datasets show that
the proposed algorithm has superior clustering performance to the
state-of-the-art clustering algorithms without any parameter
pre-specifications.
| no_new_dataset | 0.712864 |
2305.01598 | Anirudh Khatry | Anirudh Khatry, Joyce Cahoon, Jordan Henkel, Shaleen Deep, Venkatesh
Emani, Avrilia Floratou, Sumit Gulwani, Vu Le, Mohammad Raza, Sherry Shi,
Mukul Singh, Ashish Tiwari | From Words to Code: Harnessing Data for Program Synthesis from Natural
Language | 14 pages | null | null | null | cs.DB cs.AI cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Creating programs to correctly manipulate data is a difficult task, as the
underlying programming languages and APIs can be challenging to learn for many
users who are not skilled programmers. Large language models (LLMs) demonstrate
remarkable potential for generating code from natural language, but in the data
manipulation domain, apart from the natural language (NL) description of the
intended task, we also have the dataset on which the task is to be performed,
or the "data context". Existing approaches have utilized data context in a
limited way by simply adding relevant information from the input data into the
prompts sent to the LLM.
In this work, we utilize the available input data to execute the candidate
programs generated by the LLMs and gather their outputs. We introduce semantic
reranking, a technique to rerank the programs generated by LLMs based on three
signals coming the program outputs: (a) semantic filtering and well-formedness
based score tuning: do programs even generate well-formed outputs, (b) semantic
interleaving: how do the outputs from different candidates compare to each
other, and (c) output-based score tuning: how do the outputs compare to outputs
predicted for the same task. We provide theoretical justification for semantic
interleaving. We also introduce temperature mixing, where we combine samples
generated by LLMs using both high and low temperatures. We extensively evaluate
our approach in three domains, namely databases (SQL), data science (Pandas)
and business intelligence (Excel's Power Query M) on a variety of new and
existing benchmarks. We observe substantial gains across domains, with
improvements of up to 45% in top-1 accuracy and 34% in top-3 accuracy.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 16:56:32 GMT"
},
{
"version": "v2",
"created": "Wed, 3 May 2023 07:02:57 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Khatry",
"Anirudh",
""
],
[
"Cahoon",
"Joyce",
""
],
[
"Henkel",
"Jordan",
""
],
[
"Deep",
"Shaleen",
""
],
[
"Emani",
"Venkatesh",
""
],
[
"Floratou",
"Avrilia",
""
],
[
"Gulwani",
"Sumit",
""
],
[
"Le",
"Vu",
""
],
[
"Raza",
"Mohammad",
""
],
[
"Shi",
"Sherry",
""
],
[
"Singh",
"Mukul",
""
],
[
"Tiwari",
"Ashish",
""
]
] | TITLE: From Words to Code: Harnessing Data for Program Synthesis from Natural
Language
ABSTRACT: Creating programs to correctly manipulate data is a difficult task, as the
underlying programming languages and APIs can be challenging to learn for many
users who are not skilled programmers. Large language models (LLMs) demonstrate
remarkable potential for generating code from natural language, but in the data
manipulation domain, apart from the natural language (NL) description of the
intended task, we also have the dataset on which the task is to be performed,
or the "data context". Existing approaches have utilized data context in a
limited way by simply adding relevant information from the input data into the
prompts sent to the LLM.
In this work, we utilize the available input data to execute the candidate
programs generated by the LLMs and gather their outputs. We introduce semantic
reranking, a technique to rerank the programs generated by LLMs based on three
signals coming the program outputs: (a) semantic filtering and well-formedness
based score tuning: do programs even generate well-formed outputs, (b) semantic
interleaving: how do the outputs from different candidates compare to each
other, and (c) output-based score tuning: how do the outputs compare to outputs
predicted for the same task. We provide theoretical justification for semantic
interleaving. We also introduce temperature mixing, where we combine samples
generated by LLMs using both high and low temperatures. We extensively evaluate
our approach in three domains, namely databases (SQL), data science (Pandas)
and business intelligence (Excel's Power Query M) on a variety of new and
existing benchmarks. We observe substantial gains across domains, with
improvements of up to 45% in top-1 accuracy and 34% in top-3 accuracy.
| no_new_dataset | 0.711644 |
2305.01655 | Harish Pal Chauhan | Harish Chauhan, Nikunj Gupta, Zoe Haskell-Craig | Predicting blood pressure under circumstances of missing data: An
analysis of missing data patterns and imputation methods using NHANES | null | null | null | null | cs.LG stat.ME | http://creativecommons.org/publicdomain/zero/1.0/ | The World Health Organization defines cardio-vascular disease (CVD) as "a
group of disorders of the heart and blood vessels," including coronary heart
disease and stroke (WHO 21). CVD is affected by "intermediate risk factors"
such as raised blood pressure, raised blood glucose, raised blood lipids, and
obesity. These are predominantly influenced by lifestyle and behaviour,
including physical inactivity, unhealthy diets, high intake of salt, and
tobacco and alcohol use. However, genetics and social/environmental factors
such as poverty, stress, and racism also play an important role. Researchers
studying the behavioural and environmental factors associated with these
"intermediate risk factors" need access to high quality and detailed
information on diet and physical activity. However, missing data are a
pervasive problem in clinical and public health research, affecting both
randomized trials and observational studies. Reasons for missing data can vary
substantially across studies because of loss to follow-up, missed study visits,
refusal to answer survey questions, or an unrecorded measurement during an
office visit. One method of handling missing values is to simply delete
observations for which there is missingness (called Complete Case Analysis).
This is rarely used as deleting the data point containing missing data (List
wise deletion) results in a smaller number of samples and thus affects
accuracy. Additional methods of handling missing data exists, such as
summarizing the variables with its observed values (Available Case Analysis).
Motivated by the pervasiveness of missing data in the NHANES dataset, we will
conduct an analysis of imputation methods under different simulated patterns of
missing data. We will then apply these imputation methods to create a complete
dataset upon which we can use ordinary least squares to predict blood pressure
from diet and physical activity.
| [
{
"version": "v1",
"created": "Mon, 1 May 2023 18:15:44 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Chauhan",
"Harish",
""
],
[
"Gupta",
"Nikunj",
""
],
[
"Haskell-Craig",
"Zoe",
""
]
] | TITLE: Predicting blood pressure under circumstances of missing data: An
analysis of missing data patterns and imputation methods using NHANES
ABSTRACT: The World Health Organization defines cardio-vascular disease (CVD) as "a
group of disorders of the heart and blood vessels," including coronary heart
disease and stroke (WHO 21). CVD is affected by "intermediate risk factors"
such as raised blood pressure, raised blood glucose, raised blood lipids, and
obesity. These are predominantly influenced by lifestyle and behaviour,
including physical inactivity, unhealthy diets, high intake of salt, and
tobacco and alcohol use. However, genetics and social/environmental factors
such as poverty, stress, and racism also play an important role. Researchers
studying the behavioural and environmental factors associated with these
"intermediate risk factors" need access to high quality and detailed
information on diet and physical activity. However, missing data are a
pervasive problem in clinical and public health research, affecting both
randomized trials and observational studies. Reasons for missing data can vary
substantially across studies because of loss to follow-up, missed study visits,
refusal to answer survey questions, or an unrecorded measurement during an
office visit. One method of handling missing values is to simply delete
observations for which there is missingness (called Complete Case Analysis).
This is rarely used as deleting the data point containing missing data (List
wise deletion) results in a smaller number of samples and thus affects
accuracy. Additional methods of handling missing data exists, such as
summarizing the variables with its observed values (Available Case Analysis).
Motivated by the pervasiveness of missing data in the NHANES dataset, we will
conduct an analysis of imputation methods under different simulated patterns of
missing data. We will then apply these imputation methods to create a complete
dataset upon which we can use ordinary least squares to predict blood pressure
from diet and physical activity.
| no_new_dataset | 0.702913 |
2305.01663 | Prayag Tiwari Dr. | Manish Bhatia, Balram Meena, Vipin Kumar Rathi, Prayag Tiwari, Amit
Kumar Jaiswal, Shagaf M Ansari, Ajay Kumar, Pekka Marttinen | A Novel Deep Learning based Model for Erythrocytes Classification and
Quantification in Sickle Cell Disease | null | null | null | null | q-bio.QM cs.LG eess.IV | http://creativecommons.org/licenses/by/4.0/ | The shape of erythrocytes or red blood cells is altered in several
pathological conditions. Therefore, identifying and quantifying different
erythrocyte shapes can help diagnose various diseases and assist in designing a
treatment strategy. Machine Learning (ML) can be efficiently used to identify
and quantify distorted erythrocyte morphologies. In this paper, we proposed a
customized deep convolutional neural network (CNN) model to classify and
quantify the distorted and normal morphology of erythrocytes from the images
taken from the blood samples of patients suffering from Sickle cell disease (
SCD). We chose SCD as a model disease condition due to the presence of diverse
erythrocyte morphologies in the blood samples of SCD patients. For the
analysis, we used 428 raw microscopic images of SCD blood samples and generated
the dataset consisting of 10, 377 single-cell images. We focused on three
well-defined erythrocyte shapes, including discocytes, oval, and sickle. We
used 18 layered deep CNN architecture to identify and quantify these shapes
with 81% accuracy, outperforming other models. We also used SHAP and LIME for
further interpretability. The proposed model can be helpful for the quick and
accurate analysis of SCD blood samples by the clinicians and help them make the
right decision for better management of SCD.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 10:28:07 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Bhatia",
"Manish",
""
],
[
"Meena",
"Balram",
""
],
[
"Rathi",
"Vipin Kumar",
""
],
[
"Tiwari",
"Prayag",
""
],
[
"Jaiswal",
"Amit Kumar",
""
],
[
"Ansari",
"Shagaf M",
""
],
[
"Kumar",
"Ajay",
""
],
[
"Marttinen",
"Pekka",
""
]
] | TITLE: A Novel Deep Learning based Model for Erythrocytes Classification and
Quantification in Sickle Cell Disease
ABSTRACT: The shape of erythrocytes or red blood cells is altered in several
pathological conditions. Therefore, identifying and quantifying different
erythrocyte shapes can help diagnose various diseases and assist in designing a
treatment strategy. Machine Learning (ML) can be efficiently used to identify
and quantify distorted erythrocyte morphologies. In this paper, we proposed a
customized deep convolutional neural network (CNN) model to classify and
quantify the distorted and normal morphology of erythrocytes from the images
taken from the blood samples of patients suffering from Sickle cell disease (
SCD). We chose SCD as a model disease condition due to the presence of diverse
erythrocyte morphologies in the blood samples of SCD patients. For the
analysis, we used 428 raw microscopic images of SCD blood samples and generated
the dataset consisting of 10, 377 single-cell images. We focused on three
well-defined erythrocyte shapes, including discocytes, oval, and sickle. We
used 18 layered deep CNN architecture to identify and quantify these shapes
with 81% accuracy, outperforming other models. We also used SHAP and LIME for
further interpretability. The proposed model can be helpful for the quick and
accurate analysis of SCD blood samples by the clinicians and help them make the
right decision for better management of SCD.
| new_dataset | 0.710415 |
2305.01667 | Ke Zhang | Ke Zhang | Predict NAS Multi-Task by Stacking Ensemble Models using GP-NAS | Ranked 1st in CVPR 2022 Track 2 Challenge, GP-NAS, Stacking Model,
Ensemble Model | null | null | null | cs.LG cs.CV stat.AP stat.CO | http://creativecommons.org/licenses/by/4.0/ | Accurately predicting the performance of architecture with small sample
training is an important but not easy task. How to analysis and train dataset
to overcome overfitting is the core problem we should deal with. Meanwhile if
there is the mult-task problem, we should also think about if we can take
advantage of their correlation and estimate as fast as we can. In this track,
Super Network builds a search space based on ViT-Base. The search space contain
depth, num-heads, mpl-ratio and embed-dim. What we done firstly are
pre-processing the data based on our understanding of this problem which can
reduce complexity of problem and probability of over fitting. Then we tried
different kind of models and different way to combine them. Finally we choose
stacking ensemble models using GP-NAS with cross validation. Our stacking model
ranked 1st in CVPR 2022 Track 2 Challenge.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 13:59:58 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Zhang",
"Ke",
""
]
] | TITLE: Predict NAS Multi-Task by Stacking Ensemble Models using GP-NAS
ABSTRACT: Accurately predicting the performance of architecture with small sample
training is an important but not easy task. How to analysis and train dataset
to overcome overfitting is the core problem we should deal with. Meanwhile if
there is the mult-task problem, we should also think about if we can take
advantage of their correlation and estimate as fast as we can. In this track,
Super Network builds a search space based on ViT-Base. The search space contain
depth, num-heads, mpl-ratio and embed-dim. What we done firstly are
pre-processing the data based on our understanding of this problem which can
reduce complexity of problem and probability of over fitting. Then we tried
different kind of models and different way to combine them. Finally we choose
stacking ensemble models using GP-NAS with cross validation. Our stacking model
ranked 1st in CVPR 2022 Track 2 Challenge.
| no_new_dataset | 0.709252 |
2305.01668 | Xin Hong | Xin Hong, Yanyan Lan, Liang Pang, Jiafeng Guo, Xueqi Cheng | Visual Reasoning: from State to Transformation | Accepted by TPAMI. arXiv admin note: substantial text overlap with
arXiv:2011.13160 | null | 10.1109/TPAMI.2023.3268093 | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Most existing visual reasoning tasks, such as CLEVR in VQA, ignore an
important factor, i.e.~transformation. They are solely defined to test how well
machines understand concepts and relations within static settings, like one
image. Such \textbf{state driven} visual reasoning has limitations in
reflecting the ability to infer the dynamics between different states, which
has shown to be equally important for human cognition in Piaget's theory. To
tackle this problem, we propose a novel \textbf{transformation driven} visual
reasoning (TVR) task. Given both the initial and final states, the target
becomes to infer the corresponding intermediate transformation. Following this
definition, a new synthetic dataset namely TRANCE is first constructed on the
basis of CLEVR, including three levels of settings, i.e.~Basic (single-step
transformation), Event (multi-step transformation), and View (multi-step
transformation with variant views). Next, we build another real dataset called
TRANCO based on COIN, to cover the loss of transformation diversity on TRANCE.
Inspired by human reasoning, we propose a three-staged reasoning framework
called TranNet, including observing, analyzing, and concluding, to test how
recent advanced techniques perform on TVR. Experimental results show that the
state-of-the-art visual reasoning models perform well on Basic, but are still
far from human-level intelligence on Event, View, and TRANCO. We believe the
proposed new paradigm will boost the development of machine visual reasoning.
More advanced methods and new problems need to be investigated in this
direction. The resource of TVR is available at
\url{https://hongxin2019.github.io/TVR/}.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 14:24:12 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Hong",
"Xin",
""
],
[
"Lan",
"Yanyan",
""
],
[
"Pang",
"Liang",
""
],
[
"Guo",
"Jiafeng",
""
],
[
"Cheng",
"Xueqi",
""
]
] | TITLE: Visual Reasoning: from State to Transformation
ABSTRACT: Most existing visual reasoning tasks, such as CLEVR in VQA, ignore an
important factor, i.e.~transformation. They are solely defined to test how well
machines understand concepts and relations within static settings, like one
image. Such \textbf{state driven} visual reasoning has limitations in
reflecting the ability to infer the dynamics between different states, which
has shown to be equally important for human cognition in Piaget's theory. To
tackle this problem, we propose a novel \textbf{transformation driven} visual
reasoning (TVR) task. Given both the initial and final states, the target
becomes to infer the corresponding intermediate transformation. Following this
definition, a new synthetic dataset namely TRANCE is first constructed on the
basis of CLEVR, including three levels of settings, i.e.~Basic (single-step
transformation), Event (multi-step transformation), and View (multi-step
transformation with variant views). Next, we build another real dataset called
TRANCO based on COIN, to cover the loss of transformation diversity on TRANCE.
Inspired by human reasoning, we propose a three-staged reasoning framework
called TranNet, including observing, analyzing, and concluding, to test how
recent advanced techniques perform on TVR. Experimental results show that the
state-of-the-art visual reasoning models perform well on Basic, but are still
far from human-level intelligence on Event, View, and TRANCO. We believe the
proposed new paradigm will boost the development of machine visual reasoning.
More advanced methods and new problems need to be investigated in this
direction. The resource of TVR is available at
\url{https://hongxin2019.github.io/TVR/}.
| new_dataset | 0.716448 |
2305.01732 | Sunil Jaiswal | Aakash Rajpal, Noshaba Cheema, Klaus Illgner-Fehns, Philipp Slusallek,
Sunil Jaiswal | High-Resolution Synthetic RGB-D Datasets for Monocular Depth Estimation | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Accurate depth maps are essential in various applications, such as autonomous
driving, scene reconstruction, point-cloud creation, etc. However,
monocular-depth estimation (MDE) algorithms often fail to provide enough
texture & sharpness, and also are inconsistent for homogeneous scenes. These
algorithms mostly use CNN or vision transformer-based architectures requiring
large datasets for supervised training. But, MDE algorithms trained on
available depth datasets do not generalize well and hence fail to perform
accurately in diverse real-world scenes. Moreover, the ground-truth depth maps
are either lower resolution or sparse leading to relatively inconsistent depth
maps. In general, acquiring a high-resolution ground truth dataset with
pixel-level precision for accurate depth prediction is an expensive, and
time-consuming challenge.
In this paper, we generate a high-resolution synthetic depth dataset (HRSD)
of dimension 1920 X 1080 from Grand Theft Auto (GTA-V), which contains 100,000
color images and corresponding dense ground truth depth maps. The generated
datasets are diverse and have scenes from indoors to outdoors, from homogeneous
surfaces to textures. For experiments and analysis, we train the DPT algorithm,
a state-of-the-art transformer-based MDE algorithm on the proposed synthetic
dataset, which significantly increases the accuracy of depth maps on different
scenes by 9 %. Since the synthetic datasets are of higher resolution, we
propose adding a feature extraction module in the transformer encoder and
incorporating an attention-based loss, further improving the accuracy by 15 %.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 19:03:08 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Rajpal",
"Aakash",
""
],
[
"Cheema",
"Noshaba",
""
],
[
"Illgner-Fehns",
"Klaus",
""
],
[
"Slusallek",
"Philipp",
""
],
[
"Jaiswal",
"Sunil",
""
]
] | TITLE: High-Resolution Synthetic RGB-D Datasets for Monocular Depth Estimation
ABSTRACT: Accurate depth maps are essential in various applications, such as autonomous
driving, scene reconstruction, point-cloud creation, etc. However,
monocular-depth estimation (MDE) algorithms often fail to provide enough
texture & sharpness, and also are inconsistent for homogeneous scenes. These
algorithms mostly use CNN or vision transformer-based architectures requiring
large datasets for supervised training. But, MDE algorithms trained on
available depth datasets do not generalize well and hence fail to perform
accurately in diverse real-world scenes. Moreover, the ground-truth depth maps
are either lower resolution or sparse leading to relatively inconsistent depth
maps. In general, acquiring a high-resolution ground truth dataset with
pixel-level precision for accurate depth prediction is an expensive, and
time-consuming challenge.
In this paper, we generate a high-resolution synthetic depth dataset (HRSD)
of dimension 1920 X 1080 from Grand Theft Auto (GTA-V), which contains 100,000
color images and corresponding dense ground truth depth maps. The generated
datasets are diverse and have scenes from indoors to outdoors, from homogeneous
surfaces to textures. For experiments and analysis, we train the DPT algorithm,
a state-of-the-art transformer-based MDE algorithm on the proposed synthetic
dataset, which significantly increases the accuracy of depth maps on different
scenes by 9 %. Since the synthetic datasets are of higher resolution, we
propose adding a feature extraction module in the transformer encoder and
incorporating an attention-based loss, further improving the accuracy by 15 %.
| new_dataset | 0.719584 |
2305.01733 | Yuexi Zhang | Yuexi Zhang, Dan Luo, Balaji Sundareshan, Octavia Camps, Mario Sznaier | Cross-view Action Recognition via Contrastive View-invariant
Representation | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Cross view action recognition (CVAR) seeks to recognize a human action when
observed from a previously unseen viewpoint. This is a challenging problem
since the appearance of an action changes significantly with the viewpoint.
Applications of CVAR include surveillance and monitoring of assisted living
facilities where is not practical or feasible to collect large amounts of
training data when adding a new camera. We present a simple yet efficient CVAR
framework to learn invariant features from either RGB videos, 3D skeleton data,
or both. The proposed approach outperforms the current state-of-the-art
achieving similar levels of performance across input modalities: 99.4% (RGB)
and 99.9% (3D skeletons), 99.4% (RGB) and 99.9% (3D Skeletons), 97.3% (RGB),
and 99.2% (3D skeletons), and 84.4%(RGB) for the N-UCLA, NTU-RGB+D 60,
NTU-RGB+D 120, and UWA3DII datasets, respectively.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 19:04:29 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Zhang",
"Yuexi",
""
],
[
"Luo",
"Dan",
""
],
[
"Sundareshan",
"Balaji",
""
],
[
"Camps",
"Octavia",
""
],
[
"Sznaier",
"Mario",
""
]
] | TITLE: Cross-view Action Recognition via Contrastive View-invariant
Representation
ABSTRACT: Cross view action recognition (CVAR) seeks to recognize a human action when
observed from a previously unseen viewpoint. This is a challenging problem
since the appearance of an action changes significantly with the viewpoint.
Applications of CVAR include surveillance and monitoring of assisted living
facilities where is not practical or feasible to collect large amounts of
training data when adding a new camera. We present a simple yet efficient CVAR
framework to learn invariant features from either RGB videos, 3D skeleton data,
or both. The proposed approach outperforms the current state-of-the-art
achieving similar levels of performance across input modalities: 99.4% (RGB)
and 99.9% (3D skeletons), 99.4% (RGB) and 99.9% (3D Skeletons), 97.3% (RGB),
and 99.2% (3D skeletons), and 84.4%(RGB) for the N-UCLA, NTU-RGB+D 60,
NTU-RGB+D 120, and UWA3DII datasets, respectively.
| no_new_dataset | 0.709982 |
2305.01738 | Shengpu Tang | Shengpu Tang, Maggie Makar, Michael W. Sjoding, Finale Doshi-Velez,
Jenna Wiens | Leveraging Factored Action Spaces for Efficient Offline Reinforcement
Learning in Healthcare | 30 pages, 18 figures, 2 tables. NeurIPS 2022. Code available at
https://github.com/MLD3/OfflineRL_FactoredActions | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many reinforcement learning (RL) applications have combinatorial action
spaces, where each action is a composition of sub-actions. A standard RL
approach ignores this inherent factorization structure, resulting in a
potential failure to make meaningful inferences about rarely observed
sub-action combinations; this is particularly problematic for offline settings,
where data may be limited. In this work, we propose a form of linear Q-function
decomposition induced by factored action spaces. We study the theoretical
properties of our approach, identifying scenarios where it is guaranteed to
lead to zero bias when used to approximate the Q-function. Outside the regimes
with theoretical guarantees, we show that our approach can still be useful
because it leads to better sample efficiency without necessarily sacrificing
policy optimality, allowing us to achieve a better bias-variance trade-off.
Across several offline RL problems using simulators and real-world datasets
motivated by healthcare, we demonstrate that incorporating factored action
spaces into value-based RL can result in better-performing policies. Our
approach can help an agent make more accurate inferences within underexplored
regions of the state-action space when applying RL to observational datasets.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 19:13:10 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Tang",
"Shengpu",
""
],
[
"Makar",
"Maggie",
""
],
[
"Sjoding",
"Michael W.",
""
],
[
"Doshi-Velez",
"Finale",
""
],
[
"Wiens",
"Jenna",
""
]
] | TITLE: Leveraging Factored Action Spaces for Efficient Offline Reinforcement
Learning in Healthcare
ABSTRACT: Many reinforcement learning (RL) applications have combinatorial action
spaces, where each action is a composition of sub-actions. A standard RL
approach ignores this inherent factorization structure, resulting in a
potential failure to make meaningful inferences about rarely observed
sub-action combinations; this is particularly problematic for offline settings,
where data may be limited. In this work, we propose a form of linear Q-function
decomposition induced by factored action spaces. We study the theoretical
properties of our approach, identifying scenarios where it is guaranteed to
lead to zero bias when used to approximate the Q-function. Outside the regimes
with theoretical guarantees, we show that our approach can still be useful
because it leads to better sample efficiency without necessarily sacrificing
policy optimality, allowing us to achieve a better bias-variance trade-off.
Across several offline RL problems using simulators and real-world datasets
motivated by healthcare, we demonstrate that incorporating factored action
spaces into value-based RL can result in better-performing policies. Our
approach can help an agent make more accurate inferences within underexplored
regions of the state-action space when applying RL to observational datasets.
| no_new_dataset | 0.709466 |
2305.01761 | Xingbo Fu | Xingbo Fu, Chen Chen, Yushun Dong, Anil Vullikanti, Eili Klein,
Gregory Madden, Jundong Li | Spatial-Temporal Networks for Antibiogram Pattern Prediction | Accepted by the 11th IEEE International Conference on Healthcare
Informatics (IEEE ICHI 2023) | null | null | null | cs.LG cs.AI q-bio.QM | http://creativecommons.org/publicdomain/zero/1.0/ | An antibiogram is a periodic summary of antibiotic resistance results of
organisms from infected patients to selected antimicrobial drugs. Antibiograms
help clinicians to understand regional resistance rates and select appropriate
antibiotics in prescriptions. In practice, significant combinations of
antibiotic resistance may appear in different antibiograms, forming antibiogram
patterns. Such patterns may imply the prevalence of some infectious diseases in
certain regions. Thus it is of crucial importance to monitor antibiotic
resistance trends and track the spread of multi-drug resistant organisms. In
this paper, we propose a novel problem of antibiogram pattern prediction that
aims to predict which patterns will appear in the future. Despite its
importance, tackling this problem encounters a series of challenges and has not
yet been explored in the literature. First of all, antibiogram patterns are not
i.i.d as they may have strong relations with each other due to genomic
similarities of the underlying organisms. Second, antibiogram patterns are
often temporally dependent on the ones that are previously detected.
Furthermore, the spread of antibiotic resistance can be significantly
influenced by nearby or similar regions. To address the above challenges, we
propose a novel Spatial-Temporal Antibiogram Pattern Prediction framework,
STAPP, that can effectively leverage the pattern correlations and exploit the
temporal and spatial information. We conduct extensive experiments on a
real-world dataset with antibiogram reports of patients from 1999 to 2012 for
203 cities in the United States. The experimental results show the superiority
of STAPP against several competitive baselines.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 20:01:48 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Fu",
"Xingbo",
""
],
[
"Chen",
"Chen",
""
],
[
"Dong",
"Yushun",
""
],
[
"Vullikanti",
"Anil",
""
],
[
"Klein",
"Eili",
""
],
[
"Madden",
"Gregory",
""
],
[
"Li",
"Jundong",
""
]
] | TITLE: Spatial-Temporal Networks for Antibiogram Pattern Prediction
ABSTRACT: An antibiogram is a periodic summary of antibiotic resistance results of
organisms from infected patients to selected antimicrobial drugs. Antibiograms
help clinicians to understand regional resistance rates and select appropriate
antibiotics in prescriptions. In practice, significant combinations of
antibiotic resistance may appear in different antibiograms, forming antibiogram
patterns. Such patterns may imply the prevalence of some infectious diseases in
certain regions. Thus it is of crucial importance to monitor antibiotic
resistance trends and track the spread of multi-drug resistant organisms. In
this paper, we propose a novel problem of antibiogram pattern prediction that
aims to predict which patterns will appear in the future. Despite its
importance, tackling this problem encounters a series of challenges and has not
yet been explored in the literature. First of all, antibiogram patterns are not
i.i.d as they may have strong relations with each other due to genomic
similarities of the underlying organisms. Second, antibiogram patterns are
often temporally dependent on the ones that are previously detected.
Furthermore, the spread of antibiotic resistance can be significantly
influenced by nearby or similar regions. To address the above challenges, we
propose a novel Spatial-Temporal Antibiogram Pattern Prediction framework,
STAPP, that can effectively leverage the pattern correlations and exploit the
temporal and spatial information. We conduct extensive experiments on a
real-world dataset with antibiogram reports of patients from 1999 to 2012 for
203 cities in the United States. The experimental results show the superiority
of STAPP against several competitive baselines.
| no_new_dataset | 0.687617 |
2305.01764 | Zhijing Jin | Zhiheng Lyu, Zhijing Jin, Justus Mattern, Rada Mihalcea, Mrinmaya
Sachan, Bernhard Schoelkopf | Psychologically-Inspired Causal Prompts | null | null | null | null | cs.CL cs.AI cs.LG stat.ME | http://creativecommons.org/licenses/by-sa/4.0/ | NLP datasets are richer than just input-output pairs; rather, they carry
causal relations between the input and output variables. In this work, we take
sentiment classification as an example and look into the causal relations
between the review (X) and sentiment (Y). As psychology studies show that
language can affect emotion, different psychological processes are evoked when
a person first makes a rating and then self-rationalizes their feeling in a
review (where the sentiment causes the review, i.e., Y -> X), versus first
describes their experience, and weighs the pros and cons to give a final rating
(where the review causes the sentiment, i.e., X -> Y ). Furthermore, it is also
a completely different psychological process if an annotator infers the
original rating of the user by theory of mind (ToM) (where the review causes
the rating, i.e., X -ToM-> Y ). In this paper, we verbalize these three causal
mechanisms of human psychological processes of sentiment classification into
three different causal prompts, and study (1) how differently they perform, and
(2) what nature of sentiment classification data leads to agreement or
diversity in the model responses elicited by the prompts. We suggest future
work raise awareness of different causal structures in NLP tasks. Our code and
data are at https://github.com/cogito233/psych-causal-prompt
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 20:06:00 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Lyu",
"Zhiheng",
""
],
[
"Jin",
"Zhijing",
""
],
[
"Mattern",
"Justus",
""
],
[
"Mihalcea",
"Rada",
""
],
[
"Sachan",
"Mrinmaya",
""
],
[
"Schoelkopf",
"Bernhard",
""
]
] | TITLE: Psychologically-Inspired Causal Prompts
ABSTRACT: NLP datasets are richer than just input-output pairs; rather, they carry
causal relations between the input and output variables. In this work, we take
sentiment classification as an example and look into the causal relations
between the review (X) and sentiment (Y). As psychology studies show that
language can affect emotion, different psychological processes are evoked when
a person first makes a rating and then self-rationalizes their feeling in a
review (where the sentiment causes the review, i.e., Y -> X), versus first
describes their experience, and weighs the pros and cons to give a final rating
(where the review causes the sentiment, i.e., X -> Y ). Furthermore, it is also
a completely different psychological process if an annotator infers the
original rating of the user by theory of mind (ToM) (where the review causes
the rating, i.e., X -ToM-> Y ). In this paper, we verbalize these three causal
mechanisms of human psychological processes of sentiment classification into
three different causal prompts, and study (1) how differently they perform, and
(2) what nature of sentiment classification data leads to agreement or
diversity in the model responses elicited by the prompts. We suggest future
work raise awareness of different causal structures in NLP tasks. Our code and
data are at https://github.com/cogito233/psych-causal-prompt
| no_new_dataset | 0.709208 |
2305.01770 | Xinyan Li | Xinyan Li, Cheng Qian, and Lucas Glass | DeCom: Deep Coupled-Factorization Machine for Post COVID-19 Respiratory
Syncytial Virus Prediction with Nonpharmaceutical Interventions Awareness | null | null | null | null | cs.LG q-bio.QM stat.ML | http://creativecommons.org/licenses/by/4.0/ | Respiratory syncytial virus (RSV) is one of the most dangerous respiratory
diseases for infants and young children. Due to the nonpharmaceutical
intervention (NPI) imposed in the COVID-19 outbreak, the seasonal transmission
pattern of RSV has been discontinued in 2020 and then shifted months ahead in
2021 in the northern hemisphere. It is critical to understand how COVID-19
impacts RSV and build predictive algorithms to forecast the timing and
intensity of RSV reemergence in post-COVID-19 seasons. In this paper, we
propose a deep coupled tensor factorization machine, dubbed as DeCom, for post
COVID-19 RSV prediction. DeCom leverages tensor factorization and residual
modeling. It enables us to learn the disrupted RSV transmission reliably under
COVID-19 by taking both the regular seasonal RSV transmission pattern and the
NPI into consideration. Experimental results on a real RSV dataset show that
DeCom is more accurate than the state-of-the-art RSV prediction algorithms and
achieves up to 46% lower root mean square error and 49% lower mean absolute
error for country-level prediction compared to the baselines.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 20:23:09 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Li",
"Xinyan",
""
],
[
"Qian",
"Cheng",
""
],
[
"Glass",
"Lucas",
""
]
] | TITLE: DeCom: Deep Coupled-Factorization Machine for Post COVID-19 Respiratory
Syncytial Virus Prediction with Nonpharmaceutical Interventions Awareness
ABSTRACT: Respiratory syncytial virus (RSV) is one of the most dangerous respiratory
diseases for infants and young children. Due to the nonpharmaceutical
intervention (NPI) imposed in the COVID-19 outbreak, the seasonal transmission
pattern of RSV has been discontinued in 2020 and then shifted months ahead in
2021 in the northern hemisphere. It is critical to understand how COVID-19
impacts RSV and build predictive algorithms to forecast the timing and
intensity of RSV reemergence in post-COVID-19 seasons. In this paper, we
propose a deep coupled tensor factorization machine, dubbed as DeCom, for post
COVID-19 RSV prediction. DeCom leverages tensor factorization and residual
modeling. It enables us to learn the disrupted RSV transmission reliably under
COVID-19 by taking both the regular seasonal RSV transmission pattern and the
NPI into consideration. Experimental results on a real RSV dataset show that
DeCom is more accurate than the state-of-the-art RSV prediction algorithms and
achieves up to 46% lower root mean square error and 49% lower mean absolute
error for country-level prediction compared to the baselines.
| no_new_dataset | 0.710998 |
2305.01773 | Andreas Look | Andreas Look, Melih Kandemir, Barbara Rakitsch, Jan Peters | Cheap and Deterministic Inference for Deep State-Space Models of
Interacting Dynamical Systems | null | null | null | null | cs.LG cs.AI stat.ML | http://creativecommons.org/licenses/by/4.0/ | Graph neural networks are often used to model interacting dynamical systems
since they gracefully scale to systems with a varying and high number of
agents. While there has been much progress made for deterministic interacting
systems, modeling is much more challenging for stochastic systems in which one
is interested in obtaining a predictive distribution over future trajectories.
Existing methods are either computationally slow since they rely on Monte Carlo
sampling or make simplifying assumptions such that the predictive distribution
is unimodal. In this work, we present a deep state-space model which employs
graph neural networks in order to model the underlying interacting dynamical
system. The predictive distribution is multimodal and has the form of a
Gaussian mixture model, where the moments of the Gaussian components can be
computed via deterministic moment matching rules. Our moment matching scheme
can be exploited for sample-free inference, leading to more efficient and
stable training compared to Monte Carlo alternatives. Furthermore, we propose
structured approximations to the covariance matrices of the Gaussian components
in order to scale up to systems with many agents. We benchmark our novel
framework on two challenging autonomous driving datasets. Both confirm the
benefits of our method compared to state-of-the-art methods. We further
demonstrate the usefulness of our individual contributions in a carefully
designed ablation study and provide a detailed runtime analysis of our proposed
covariance approximations. Finally, we empirically demonstrate the
generalization ability of our method by evaluating its performance on unseen
scenarios.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 20:30:23 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Look",
"Andreas",
""
],
[
"Kandemir",
"Melih",
""
],
[
"Rakitsch",
"Barbara",
""
],
[
"Peters",
"Jan",
""
]
] | TITLE: Cheap and Deterministic Inference for Deep State-Space Models of
Interacting Dynamical Systems
ABSTRACT: Graph neural networks are often used to model interacting dynamical systems
since they gracefully scale to systems with a varying and high number of
agents. While there has been much progress made for deterministic interacting
systems, modeling is much more challenging for stochastic systems in which one
is interested in obtaining a predictive distribution over future trajectories.
Existing methods are either computationally slow since they rely on Monte Carlo
sampling or make simplifying assumptions such that the predictive distribution
is unimodal. In this work, we present a deep state-space model which employs
graph neural networks in order to model the underlying interacting dynamical
system. The predictive distribution is multimodal and has the form of a
Gaussian mixture model, where the moments of the Gaussian components can be
computed via deterministic moment matching rules. Our moment matching scheme
can be exploited for sample-free inference, leading to more efficient and
stable training compared to Monte Carlo alternatives. Furthermore, we propose
structured approximations to the covariance matrices of the Gaussian components
in order to scale up to systems with many agents. We benchmark our novel
framework on two challenging autonomous driving datasets. Both confirm the
benefits of our method compared to state-of-the-art methods. We further
demonstrate the usefulness of our individual contributions in a carefully
designed ablation study and provide a detailed runtime analysis of our proposed
covariance approximations. Finally, we empirically demonstrate the
generalization ability of our method by evaluating its performance on unseen
scenarios.
| no_new_dataset | 0.708811 |
2305.01795 | Yujie Lu | Yujie Lu, Pan Lu, Zhiyu Chen, Wanrong Zhu, Xin Eric Wang, William Yang
Wang | Multimodal Procedural Planning via Dual Text-Image Prompting | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Embodied agents have achieved prominent performance in following human
instructions to complete tasks. However, the potential of providing
instructions informed by texts and images to assist humans in completing tasks
remains underexplored. To uncover this capability, we present the multimodal
procedural planning (MPP) task, in which models are given a high-level goal and
generate plans of paired text-image steps, providing more complementary and
informative guidance than unimodal plans. The key challenges of MPP are to
ensure the informativeness, temporal coherence,and accuracy of plans across
modalities. To tackle this, we propose Text-Image Prompting (TIP), a
dual-modality prompting method that jointly leverages zero-shot reasoning
ability in large language models (LLMs) and compelling text-to-image generation
ability from diffusion-based models. TIP improves the interaction in the dual
modalities using Text-to-Image Bridge and Image-to-Text Bridge, allowing LLMs
to guide the textual-grounded image plan generation and leveraging the
descriptions of image plans to ground the textual plan reversely. To address
the lack of relevant datasets, we collect WIKIPLAN and RECIPEPLAN as a testbed
for MPP. Our results show compelling human preferences and automatic scores
against unimodal and multimodal baselines on WIKIPLAN and RECIPEPLAN in terms
of informativeness, temporal coherence, and plan accuracy. Our code and data:
https://github.com/YujieLu10/MPP.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 21:46:44 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Lu",
"Yujie",
""
],
[
"Lu",
"Pan",
""
],
[
"Chen",
"Zhiyu",
""
],
[
"Zhu",
"Wanrong",
""
],
[
"Wang",
"Xin Eric",
""
],
[
"Wang",
"William Yang",
""
]
] | TITLE: Multimodal Procedural Planning via Dual Text-Image Prompting
ABSTRACT: Embodied agents have achieved prominent performance in following human
instructions to complete tasks. However, the potential of providing
instructions informed by texts and images to assist humans in completing tasks
remains underexplored. To uncover this capability, we present the multimodal
procedural planning (MPP) task, in which models are given a high-level goal and
generate plans of paired text-image steps, providing more complementary and
informative guidance than unimodal plans. The key challenges of MPP are to
ensure the informativeness, temporal coherence,and accuracy of plans across
modalities. To tackle this, we propose Text-Image Prompting (TIP), a
dual-modality prompting method that jointly leverages zero-shot reasoning
ability in large language models (LLMs) and compelling text-to-image generation
ability from diffusion-based models. TIP improves the interaction in the dual
modalities using Text-to-Image Bridge and Image-to-Text Bridge, allowing LLMs
to guide the textual-grounded image plan generation and leveraging the
descriptions of image plans to ground the textual plan reversely. To address
the lack of relevant datasets, we collect WIKIPLAN and RECIPEPLAN as a testbed
for MPP. Our results show compelling human preferences and automatic scores
against unimodal and multimodal baselines on WIKIPLAN and RECIPEPLAN in terms
of informativeness, temporal coherence, and plan accuracy. Our code and data:
https://github.com/YujieLu10/MPP.
| no_new_dataset | 0.709787 |
2305.01801 | Yushun Dong | Yushun Dong, Jundong Li, Tobias Schnabel | When Newer is Not Better: Does Deep Learning Really Benefit
Recommendation From Implicit Feedback? | Published as a conference paper at SIGIR 2023 | null | null | null | cs.IR cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, neural models have been repeatedly touted to exhibit
state-of-the-art performance in recommendation. Nevertheless, multiple recent
studies have revealed that the reported state-of-the-art results of many neural
recommendation models cannot be reliably replicated. A primary reason is that
existing evaluations are performed under various inconsistent protocols.
Correspondingly, these replicability issues make it difficult to understand how
much benefit we can actually gain from these neural models. It then becomes
clear that a fair and comprehensive performance comparison between traditional
and neural models is needed.
Motivated by these issues, we perform a large-scale, systematic study to
compare recent neural recommendation models against traditional ones in top-n
recommendation from implicit data. We propose a set of evaluation strategies
for measuring memorization performance, generalization performance, and
subgroup-specific performance of recommendation models. We conduct extensive
experiments with 13 popular recommendation models (including two neural models
and 11 traditional ones as baselines) on nine commonly used datasets. Our
experiments demonstrate that even with extensive hyper-parameter searches,
neural models do not dominate traditional models in all aspects, e.g., they
fare worse in terms of average HitRate. We further find that there are areas
where neural models seem to outperform non-neural models, for example, in
recommendation diversity and robustness between different subgroups of users
and items. Our work illuminates the relative advantages and disadvantages of
neural models in recommendation and is therefore an important step towards
building better recommender systems.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 22:03:49 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Dong",
"Yushun",
""
],
[
"Li",
"Jundong",
""
],
[
"Schnabel",
"Tobias",
""
]
] | TITLE: When Newer is Not Better: Does Deep Learning Really Benefit
Recommendation From Implicit Feedback?
ABSTRACT: In recent years, neural models have been repeatedly touted to exhibit
state-of-the-art performance in recommendation. Nevertheless, multiple recent
studies have revealed that the reported state-of-the-art results of many neural
recommendation models cannot be reliably replicated. A primary reason is that
existing evaluations are performed under various inconsistent protocols.
Correspondingly, these replicability issues make it difficult to understand how
much benefit we can actually gain from these neural models. It then becomes
clear that a fair and comprehensive performance comparison between traditional
and neural models is needed.
Motivated by these issues, we perform a large-scale, systematic study to
compare recent neural recommendation models against traditional ones in top-n
recommendation from implicit data. We propose a set of evaluation strategies
for measuring memorization performance, generalization performance, and
subgroup-specific performance of recommendation models. We conduct extensive
experiments with 13 popular recommendation models (including two neural models
and 11 traditional ones as baselines) on nine commonly used datasets. Our
experiments demonstrate that even with extensive hyper-parameter searches,
neural models do not dominate traditional models in all aspects, e.g., they
fare worse in terms of average HitRate. We further find that there are areas
where neural models seem to outperform non-neural models, for example, in
recommendation diversity and robustness between different subgroups of users
and items. Our work illuminates the relative advantages and disadvantages of
neural models in recommendation and is therefore an important step towards
building better recommender systems.
| no_new_dataset | 0.706431 |
2305.01812 | Neeraj Varshney | Neeraj Varshney and Chitta Baral | Post-Abstention: Towards Reliably Re-Attempting the Abstained Instances
in QA | ACL 2023 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite remarkable progress made in natural language processing, even the
state-of-the-art models often make incorrect predictions. Such predictions
hamper the reliability of systems and limit their widespread adoption in
real-world applications. 'Selective prediction' partly addresses the above
concern by enabling models to abstain from answering when their predictions are
likely to be incorrect. While selective prediction is advantageous, it leaves
us with a pertinent question 'what to do after abstention'. To this end, we
present an explorative study on 'Post-Abstention', a task that allows
re-attempting the abstained instances with the aim of increasing 'coverage' of
the system without significantly sacrificing its 'accuracy'. We first provide
mathematical formulation of this task and then explore several methods to solve
it. Comprehensive experiments on 11 QA datasets show that these methods lead to
considerable risk improvements -- performance metric of the Post-Abstention
task -- both in the in-domain and the out-of-domain settings. We also conduct a
thorough analysis of these results which further leads to several interesting
findings. Finally, we believe that our work will encourage and facilitate
further research in this important area of addressing the reliability of NLP
systems.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 22:40:41 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Varshney",
"Neeraj",
""
],
[
"Baral",
"Chitta",
""
]
] | TITLE: Post-Abstention: Towards Reliably Re-Attempting the Abstained Instances
in QA
ABSTRACT: Despite remarkable progress made in natural language processing, even the
state-of-the-art models often make incorrect predictions. Such predictions
hamper the reliability of systems and limit their widespread adoption in
real-world applications. 'Selective prediction' partly addresses the above
concern by enabling models to abstain from answering when their predictions are
likely to be incorrect. While selective prediction is advantageous, it leaves
us with a pertinent question 'what to do after abstention'. To this end, we
present an explorative study on 'Post-Abstention', a task that allows
re-attempting the abstained instances with the aim of increasing 'coverage' of
the system without significantly sacrificing its 'accuracy'. We first provide
mathematical formulation of this task and then explore several methods to solve
it. Comprehensive experiments on 11 QA datasets show that these methods lead to
considerable risk improvements -- performance metric of the Post-Abstention
task -- both in the in-domain and the out-of-domain settings. We also conduct a
thorough analysis of these results which further leads to several interesting
findings. Finally, we believe that our work will encourage and facilitate
further research in this important area of addressing the reliability of NLP
systems.
| no_new_dataset | 0.704592 |
2305.01822 | Katherine Deck | Tobias Bischoff and Katherine Deck | Unpaired Downscaling of Fluid Flows with Diffusion Bridges | Submitted to Artificial Intelligence for the Earth Systems | null | null | null | cs.LG physics.flu-dyn physics.geo-ph | http://creativecommons.org/licenses/by/4.0/ | We present a method to downscale idealized geophysical fluid simulations
using generative models based on diffusion maps. By analyzing the Fourier
spectra of images drawn from different data distributions, we show how one can
chain together two independent conditional diffusion models for use in domain
translation. The resulting transformation is a diffusion bridge between a low
resolution and a high resolution dataset and allows for new sample generation
of high-resolution images given specific low resolution features. The ability
to generate new samples allows for the computation of any statistic of
interest, without any additional calibration or training. Our unsupervised
setup is also designed to downscale images without access to paired training
data; this flexibility allows for the combination of multiple source and target
domains without additional training. We demonstrate that the method enhances
resolution and corrects context-dependent biases in geophysical fluid
simulations, including in extreme events. We anticipate that the same method
can be used to downscale the output of climate simulations, including
temperature and precipitation fields, without needing to train a new model for
each application and providing a significant computational cost savings.
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 23:13:44 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Bischoff",
"Tobias",
""
],
[
"Deck",
"Katherine",
""
]
] | TITLE: Unpaired Downscaling of Fluid Flows with Diffusion Bridges
ABSTRACT: We present a method to downscale idealized geophysical fluid simulations
using generative models based on diffusion maps. By analyzing the Fourier
spectra of images drawn from different data distributions, we show how one can
chain together two independent conditional diffusion models for use in domain
translation. The resulting transformation is a diffusion bridge between a low
resolution and a high resolution dataset and allows for new sample generation
of high-resolution images given specific low resolution features. The ability
to generate new samples allows for the computation of any statistic of
interest, without any additional calibration or training. Our unsupervised
setup is also designed to downscale images without access to paired training
data; this flexibility allows for the combination of multiple source and target
domains without additional training. We demonstrate that the method enhances
resolution and corrects context-dependent biases in geophysical fluid
simulations, including in extreme events. We anticipate that the same method
can be used to downscale the output of climate simulations, including
temperature and precipitation fields, without needing to train a new model for
each application and providing a significant computational cost savings.
| no_new_dataset | 0.712614 |
2305.01827 | Karthik Gopinath | Karthik Gopinath, Douglas N. Greve, Sudeshna Das, Steve Arnold, Colin
Magdamo, and Juan Eugenio Iglesias | Cortical analysis of heterogeneous clinical brain MRI scans for
large-scale neuroimaging studies | null | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Surface analysis of the cortex is ubiquitous in human neuroimaging with MRI,
e.g., for cortical registration, parcellation, or thickness estimation. The
convoluted cortical geometry requires isotropic scans (e.g., 1mm MPRAGEs) and
good gray-white matter contrast for 3D reconstruction. This precludes the
analysis of most brain MRI scans acquired for clinical purposes. Analyzing such
scans would enable neuroimaging studies with sample sizes that cannot be
achieved with current research datasets, particularly for underrepresented
populations and rare diseases. Here we present the first method for cortical
reconstruction, registration, parcellation, and thickness estimation for
clinical brain MRI scans of any resolution and pulse sequence. The methods has
a learning component and a classical optimization module. The former uses
domain randomization to train a CNN that predicts an implicit representation of
the white matter and pial surfaces (a signed distance function) at 1mm
isotropic resolution, independently of the pulse sequence and resolution of the
input. The latter uses geometry processing to place the surfaces while
accurately satisfying topological and geometric constraints, thus enabling
subsequent parcellation and thickness estimation with existing methods. We
present results on 5mm axial FLAIR scans from ADNI and on a highly
heterogeneous clinical dataset with 5,000 scans. Code and data are publicly
available at https://surfer.nmr.mgh.harvard.edu/fswiki/recon-all-clinical
| [
{
"version": "v1",
"created": "Tue, 2 May 2023 23:36:06 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Gopinath",
"Karthik",
""
],
[
"Greve",
"Douglas N.",
""
],
[
"Das",
"Sudeshna",
""
],
[
"Arnold",
"Steve",
""
],
[
"Magdamo",
"Colin",
""
],
[
"Iglesias",
"Juan Eugenio",
""
]
] | TITLE: Cortical analysis of heterogeneous clinical brain MRI scans for
large-scale neuroimaging studies
ABSTRACT: Surface analysis of the cortex is ubiquitous in human neuroimaging with MRI,
e.g., for cortical registration, parcellation, or thickness estimation. The
convoluted cortical geometry requires isotropic scans (e.g., 1mm MPRAGEs) and
good gray-white matter contrast for 3D reconstruction. This precludes the
analysis of most brain MRI scans acquired for clinical purposes. Analyzing such
scans would enable neuroimaging studies with sample sizes that cannot be
achieved with current research datasets, particularly for underrepresented
populations and rare diseases. Here we present the first method for cortical
reconstruction, registration, parcellation, and thickness estimation for
clinical brain MRI scans of any resolution and pulse sequence. The methods has
a learning component and a classical optimization module. The former uses
domain randomization to train a CNN that predicts an implicit representation of
the white matter and pial surfaces (a signed distance function) at 1mm
isotropic resolution, independently of the pulse sequence and resolution of the
input. The latter uses geometry processing to place the surfaces while
accurately satisfying topological and geometric constraints, thus enabling
subsequent parcellation and thickness estimation with existing methods. We
present results on 5mm axial FLAIR scans from ADNI and on a highly
heterogeneous clinical dataset with 5,000 scans. Code and data are publicly
available at https://surfer.nmr.mgh.harvard.edu/fswiki/recon-all-clinical
| no_new_dataset | 0.711844 |
2305.01836 | Shentong Mo | Shentong Mo, Yapeng Tian | AV-SAM: Segment Anything Model Meets Audio-Visual Localization and
Segmentation | null | null | null | null | cs.CV cs.LG cs.MM cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Segment Anything Model (SAM) has recently shown its powerful effectiveness in
visual segmentation tasks. However, there is less exploration concerning how
SAM works on audio-visual tasks, such as visual sound localization and
segmentation. In this work, we propose a simple yet effective audio-visual
localization and segmentation framework based on the Segment Anything Model,
namely AV-SAM, that can generate sounding object masks corresponding to the
audio. Specifically, our AV-SAM simply leverages pixel-wise audio-visual fusion
across audio features and visual features from the pre-trained image encoder in
SAM to aggregate cross-modal representations. Then, the aggregated cross-modal
features are fed into the prompt encoder and mask decoder to generate the final
audio-visual segmentation masks. We conduct extensive experiments on
Flickr-SoundNet and AVSBench datasets. The results demonstrate that the
proposed AV-SAM can achieve competitive performance on sounding object
localization and segmentation.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 00:33:52 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Mo",
"Shentong",
""
],
[
"Tian",
"Yapeng",
""
]
] | TITLE: AV-SAM: Segment Anything Model Meets Audio-Visual Localization and
Segmentation
ABSTRACT: Segment Anything Model (SAM) has recently shown its powerful effectiveness in
visual segmentation tasks. However, there is less exploration concerning how
SAM works on audio-visual tasks, such as visual sound localization and
segmentation. In this work, we propose a simple yet effective audio-visual
localization and segmentation framework based on the Segment Anything Model,
namely AV-SAM, that can generate sounding object masks corresponding to the
audio. Specifically, our AV-SAM simply leverages pixel-wise audio-visual fusion
across audio features and visual features from the pre-trained image encoder in
SAM to aggregate cross-modal representations. Then, the aggregated cross-modal
features are fed into the prompt encoder and mask decoder to generate the final
audio-visual segmentation masks. We conduct extensive experiments on
Flickr-SoundNet and AVSBench datasets. The results demonstrate that the
proposed AV-SAM can achieve competitive performance on sounding object
localization and segmentation.
| no_new_dataset | 0.710465 |
2305.01837 | Mahesh Bhosale Mr | Jay Lal, Aditya Mitkari, Mahesh Bhosale, David Doermann | LineFormer: Rethinking Line Chart Data Extraction as Instance
Segmentation | Accepted to ICDAR 2023 | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Data extraction from line-chart images is an essential component of the
automated document understanding process, as line charts are a ubiquitous data
visualization format. However, the amount of visual and structural variations
in multi-line graphs makes them particularly challenging for automated parsing.
Existing works, however, are not robust to all these variations, either taking
an all-chart unified approach or relying on auxiliary information such as
legends for line data extraction. In this work, we propose LineFormer, a robust
approach to line data extraction using instance segmentation. We achieve
state-of-the-art performance on several benchmark synthetic and real chart
datasets. Our implementation is available at
https://github.com/TheJaeLal/LineFormer .
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 00:38:24 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Lal",
"Jay",
""
],
[
"Mitkari",
"Aditya",
""
],
[
"Bhosale",
"Mahesh",
""
],
[
"Doermann",
"David",
""
]
] | TITLE: LineFormer: Rethinking Line Chart Data Extraction as Instance
Segmentation
ABSTRACT: Data extraction from line-chart images is an essential component of the
automated document understanding process, as line charts are a ubiquitous data
visualization format. However, the amount of visual and structural variations
in multi-line graphs makes them particularly challenging for automated parsing.
Existing works, however, are not robust to all these variations, either taking
an all-chart unified approach or relying on auxiliary information such as
legends for line data extraction. In this work, we propose LineFormer, a robust
approach to line data extraction using instance segmentation. We achieve
state-of-the-art performance on several benchmark synthetic and real chart
datasets. Our implementation is available at
https://github.com/TheJaeLal/LineFormer .
| no_new_dataset | 0.71287 |
2305.01843 | Kenny Chen | Kenny Chen, Ryan Nemiroff, Brett T. Lopez | Direct LiDAR-Inertial Odometry and Mapping: Perceptive and Connective
SLAM | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents Direct LiDAR-Inertial Odometry and Mapping (DLIOM), a
robust SLAM algorithm with an explicit focus on computational efficiency,
operational reliability, and real-world efficacy. DLIOM contains several key
algorithmic innovations in both the front-end and back-end subsystems to design
a resilient LiDAR-inertial architecture that is perceptive to the environment
and produces accurate localization and high-fidelity 3D mapping for autonomous
robotic platforms. Our ideas spawned after a deep investigation into modern
LiDAR SLAM systems and their inabilities to generalize across different
operating environments, in which we address several common algorithmic failure
points by means of proactive safe-guards to provide long-term operational
reliability in the unstructured real world. We detail several important
innovations to localization accuracy and mapping resiliency distributed
throughout a typical LiDAR SLAM pipeline to comprehensively increase
algorithmic speed, accuracy, and robustness. In addition, we discuss insights
gained from our ground-up approach while implementing such a complex system for
real-time state estimation on resource-constrained systems, and we
experimentally show the increased performance of our method as compared to the
current state-of-the-art on both public benchmark and self-collected datasets.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 01:06:25 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Chen",
"Kenny",
""
],
[
"Nemiroff",
"Ryan",
""
],
[
"Lopez",
"Brett T.",
""
]
] | TITLE: Direct LiDAR-Inertial Odometry and Mapping: Perceptive and Connective
SLAM
ABSTRACT: This paper presents Direct LiDAR-Inertial Odometry and Mapping (DLIOM), a
robust SLAM algorithm with an explicit focus on computational efficiency,
operational reliability, and real-world efficacy. DLIOM contains several key
algorithmic innovations in both the front-end and back-end subsystems to design
a resilient LiDAR-inertial architecture that is perceptive to the environment
and produces accurate localization and high-fidelity 3D mapping for autonomous
robotic platforms. Our ideas spawned after a deep investigation into modern
LiDAR SLAM systems and their inabilities to generalize across different
operating environments, in which we address several common algorithmic failure
points by means of proactive safe-guards to provide long-term operational
reliability in the unstructured real world. We detail several important
innovations to localization accuracy and mapping resiliency distributed
throughout a typical LiDAR SLAM pipeline to comprehensively increase
algorithmic speed, accuracy, and robustness. In addition, we discuss insights
gained from our ground-up approach while implementing such a complex system for
real-time state estimation on resource-constrained systems, and we
experimentally show the increased performance of our method as compared to the
current state-of-the-art on both public benchmark and self-collected datasets.
| no_new_dataset | 0.709265 |
2305.01868 | Daochen Zha | Daochen Zha, Louis Feng, Liang Luo, Bhargav Bhushanam, Zirui Liu,
Yusuo Hu, Jade Nie, Yuzhen Huang, Yuandong Tian, Arun Kejariwal, Xia Hu | Pre-train and Search: Efficient Embedding Table Sharding with
Pre-trained Neural Cost Models | Accepted by MLSys 2023. Code available at
https://github.com/daochenzha/neuroshard | null | null | null | cs.LG cs.DC cs.IR cs.PF | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sharding a large machine learning model across multiple devices to balance
the costs is important in distributed training. This is challenging because
partitioning is NP-hard, and estimating the costs accurately and efficiently is
difficult. In this work, we explore a "pre-train, and search" paradigm for
efficient sharding. The idea is to pre-train a universal and once-for-all
neural network to predict the costs of all the possible shards, which serves as
an efficient sharding simulator. Built upon this pre-trained cost model, we
then perform an online search to identify the best sharding plans given any
specific sharding task. We instantiate this idea in deep learning
recommendation models (DLRMs) and propose NeuroShard for embedding table
sharding. NeuroShard pre-trains neural cost models on augmented tables to cover
various sharding scenarios. Then it identifies the best column-wise and
table-wise sharding plans with beam search and greedy grid search,
respectively. Experiments show that NeuroShard significantly and consistently
outperforms the state-of-the-art on the benchmark sharding dataset, achieving
up to 23.8% improvement. When deployed in an ultra-large production DLRM with
multi-terabyte embedding tables, NeuroShard achieves 11.6% improvement in
embedding costs over the state-of-the-art, which translates to 6.6% end-to-end
training throughput improvement. To facilitate future research of the
"pre-train, and search" paradigm in ML for Systems, we open-source our code at
https://github.com/daochenzha/neuroshard
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 02:52:03 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Zha",
"Daochen",
""
],
[
"Feng",
"Louis",
""
],
[
"Luo",
"Liang",
""
],
[
"Bhushanam",
"Bhargav",
""
],
[
"Liu",
"Zirui",
""
],
[
"Hu",
"Yusuo",
""
],
[
"Nie",
"Jade",
""
],
[
"Huang",
"Yuzhen",
""
],
[
"Tian",
"Yuandong",
""
],
[
"Kejariwal",
"Arun",
""
],
[
"Hu",
"Xia",
""
]
] | TITLE: Pre-train and Search: Efficient Embedding Table Sharding with
Pre-trained Neural Cost Models
ABSTRACT: Sharding a large machine learning model across multiple devices to balance
the costs is important in distributed training. This is challenging because
partitioning is NP-hard, and estimating the costs accurately and efficiently is
difficult. In this work, we explore a "pre-train, and search" paradigm for
efficient sharding. The idea is to pre-train a universal and once-for-all
neural network to predict the costs of all the possible shards, which serves as
an efficient sharding simulator. Built upon this pre-trained cost model, we
then perform an online search to identify the best sharding plans given any
specific sharding task. We instantiate this idea in deep learning
recommendation models (DLRMs) and propose NeuroShard for embedding table
sharding. NeuroShard pre-trains neural cost models on augmented tables to cover
various sharding scenarios. Then it identifies the best column-wise and
table-wise sharding plans with beam search and greedy grid search,
respectively. Experiments show that NeuroShard significantly and consistently
outperforms the state-of-the-art on the benchmark sharding dataset, achieving
up to 23.8% improvement. When deployed in an ultra-large production DLRM with
multi-terabyte embedding tables, NeuroShard achieves 11.6% improvement in
embedding costs over the state-of-the-art, which translates to 6.6% end-to-end
training throughput improvement. To facilitate future research of the
"pre-train, and search" paradigm in ML for Systems, we open-source our code at
https://github.com/daochenzha/neuroshard
| no_new_dataset | 0.70883 |
2305.01873 | Ruslan Isaev Dr. | Dim Shaiakhmetov, Remudin Reshid Mekuria, Ruslan Isaev and Fatma Unsal | Morphological Classification of Galaxies Using SpinalNet | 5 pages, 4 figures, ICECCO conference | D. Shaiakhmetov, R. R. Mekuria, R. Isaev and F. Unsal,
"Morphological Classification of Galaxies Using SpinalNet," 2021 16th
International Conference on Electronics Computer and Computation (ICECCO),
Kaskelen, Kazakhstan, 2021, pp. 1-5 | 10.1109/ICECCO53203.2021.9663784 | null | cs.LG cs.CV | http://creativecommons.org/licenses/by/4.0/ | Deep neural networks (DNNs) with a step-by-step introduction of inputs, which
is constructed by imitating the somatosensory system in human body, known as
SpinalNet have been implemented in this work on a Galaxy Zoo dataset. The input
segmentation in SpinalNet has enabled the intermediate layers to take some of
the inputs as well as output of preceding layers thereby reducing the amount of
the collected weights in the intermediate layers. As a result of these, the
authors of SpinalNet reported to have achieved in most of the DNNs they tested,
not only a remarkable cut in the error but also in the large reduction of the
computational costs. Having applied it to the Galaxy Zoo dataset, we are able
to classify the different classes and/or sub-classes of the galaxies. Thus, we
have obtained higher classification accuracies of 98.2, 95 and 82 percents
between elliptical and spirals, between these two and irregulars, and between
10 sub-classes of galaxies, respectively.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 03:20:18 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Shaiakhmetov",
"Dim",
""
],
[
"Mekuria",
"Remudin Reshid",
""
],
[
"Isaev",
"Ruslan",
""
],
[
"Unsal",
"Fatma",
""
]
] | TITLE: Morphological Classification of Galaxies Using SpinalNet
ABSTRACT: Deep neural networks (DNNs) with a step-by-step introduction of inputs, which
is constructed by imitating the somatosensory system in human body, known as
SpinalNet have been implemented in this work on a Galaxy Zoo dataset. The input
segmentation in SpinalNet has enabled the intermediate layers to take some of
the inputs as well as output of preceding layers thereby reducing the amount of
the collected weights in the intermediate layers. As a result of these, the
authors of SpinalNet reported to have achieved in most of the DNNs they tested,
not only a remarkable cut in the error but also in the large reduction of the
computational costs. Having applied it to the Galaxy Zoo dataset, we are able
to classify the different classes and/or sub-classes of the galaxies. Thus, we
have obtained higher classification accuracies of 98.2, 95 and 82 percents
between elliptical and spirals, between these two and irregulars, and between
10 sub-classes of galaxies, respectively.
| no_new_dataset | 0.709648 |
2305.01884 | Naveen Siva Kumar Badveeti | Darshan Gera, Badveeti Naveen Siva Kumar, Bobbili Veerendra Raj Kumar,
S Balasubramanian | Class adaptive threshold and negative class guided noisy annotation
robust Facial Expression Recognition | 14 pages, 9 figures | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | The hindering problem in facial expression recognition (FER) is the presence
of inaccurate annotations referred to as noisy annotations in the datasets.
These noisy annotations are present in the datasets inherently because the
labeling is subjective to the annotator, clarity of the image, etc. Recent
works use sample selection methods to solve this noisy annotation problem in
FER. In our work, we use a dynamic adaptive threshold to separate confident
samples from non-confident ones so that our learning won't be hampered due to
non-confident samples. Instead of discarding the non-confident samples, we
impose consistency in the negative classes of those non-confident samples to
guide the model to learn better in the positive class. Since FER datasets
usually come with 7 or 8 classes, we can correctly guess a negative class by
85% probability even by choosing randomly. By learning "which class a sample
doesn't belong to", the model can learn "which class it belongs to" in a better
manner. We demonstrate proposed framework's effectiveness using quantitative as
well as qualitative results. Our method performs better than the baseline by a
margin of 4% to 28% on RAFDB and 3.3% to 31.4% on FERPlus for various levels of
synthetic noisy labels in the aforementioned datasets.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 04:28:49 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Gera",
"Darshan",
""
],
[
"Kumar",
"Badveeti Naveen Siva",
""
],
[
"Kumar",
"Bobbili Veerendra Raj",
""
],
[
"Balasubramanian",
"S",
""
]
] | TITLE: Class adaptive threshold and negative class guided noisy annotation
robust Facial Expression Recognition
ABSTRACT: The hindering problem in facial expression recognition (FER) is the presence
of inaccurate annotations referred to as noisy annotations in the datasets.
These noisy annotations are present in the datasets inherently because the
labeling is subjective to the annotator, clarity of the image, etc. Recent
works use sample selection methods to solve this noisy annotation problem in
FER. In our work, we use a dynamic adaptive threshold to separate confident
samples from non-confident ones so that our learning won't be hampered due to
non-confident samples. Instead of discarding the non-confident samples, we
impose consistency in the negative classes of those non-confident samples to
guide the model to learn better in the positive class. Since FER datasets
usually come with 7 or 8 classes, we can correctly guess a negative class by
85% probability even by choosing randomly. By learning "which class a sample
doesn't belong to", the model can learn "which class it belongs to" in a better
manner. We demonstrate proposed framework's effectiveness using quantitative as
well as qualitative results. Our method performs better than the baseline by a
margin of 4% to 28% on RAFDB and 3.3% to 31.4% on FERPlus for various levels of
synthetic noisy labels in the aforementioned datasets.
| no_new_dataset | 0.713226 |
2305.01886 | Gargi Alavani | Gargi Alavani, Santonu Sarkar | Prediction of Performance and Power Consumption of GPGPU Applications | PhD Thesis | Gargi Alavani, Santonu Sarkar, "Performance modeling of graphics
processing unit application using static and dynamic analysis", Concurrency
and Computation: Practice and Experience (CPE), Volume 34, 2021 | null | null | cs.DC cs.PF | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Graphics Processing Units (GPUs) have become an integral part of
High-Performance Computing to achieve an Exascale performance. The main goal of
application developers of GPU is to tune their code extensively to obtain
optimal performance, making efficient use of different resources available.
While extracting optimal performance of applications on an HPC infrastructure,
developers should also ensure the applications have the least energy usage
considering the massive power consumption of data centres and HPC servers. This
thesis presents two models developed which can be utilized by developers in
analysing the CUDA kernel's energy dissipation. The first one is a model that
predicts the CUDA kernel's execution time. Here a PTX code is statically
analysed to extract instruction features, control flow, and data dependence. We
propose two scheduling algorithm approaches that satisfy the performance and
hardware constraints. The second model is a static analysis-based power
prediction built by utilizing machine learning techniques. Features used for
building the model are derived using static analysis of PTX code. These
features are chosen to understand the relationship between GPU power
consumption and program features that can aid developers in building
energy-efficient, sustainable applications. The dataset used for validating
both models include kernels from different benchmarks suits, sizes, nature
(e.g., compute-bound, memory-bound), and complexity (e.g., control divergence,
memory access patterns). We also present a tool that has practically validated
the effectiveness and ease of using the two models as design assistance tools
for GPU.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 04:31:43 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Alavani",
"Gargi",
""
],
[
"Sarkar",
"Santonu",
""
]
] | TITLE: Prediction of Performance and Power Consumption of GPGPU Applications
ABSTRACT: Graphics Processing Units (GPUs) have become an integral part of
High-Performance Computing to achieve an Exascale performance. The main goal of
application developers of GPU is to tune their code extensively to obtain
optimal performance, making efficient use of different resources available.
While extracting optimal performance of applications on an HPC infrastructure,
developers should also ensure the applications have the least energy usage
considering the massive power consumption of data centres and HPC servers. This
thesis presents two models developed which can be utilized by developers in
analysing the CUDA kernel's energy dissipation. The first one is a model that
predicts the CUDA kernel's execution time. Here a PTX code is statically
analysed to extract instruction features, control flow, and data dependence. We
propose two scheduling algorithm approaches that satisfy the performance and
hardware constraints. The second model is a static analysis-based power
prediction built by utilizing machine learning techniques. Features used for
building the model are derived using static analysis of PTX code. These
features are chosen to understand the relationship between GPU power
consumption and program features that can aid developers in building
energy-efficient, sustainable applications. The dataset used for validating
both models include kernels from different benchmarks suits, sizes, nature
(e.g., compute-bound, memory-bound), and complexity (e.g., control divergence,
memory access patterns). We also present a tool that has practically validated
the effectiveness and ease of using the two models as design assistance tools
for GPU.
| no_new_dataset | 0.708055 |
2305.01888 | Lavisha Aggarwal | Lavisha Aggarwal, Shruti Bhargava | Fairness in AI Systems: Mitigating gender bias from language-vision
models | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Our society is plagued by several biases, including racial biases, caste
biases, and gender bias. As a matter of fact, several years ago, most of these
notions were unheard of. These biases passed through generations along with
amplification have lead to scenarios where these have taken the role of
expected norms by certain groups in the society. One notable example is of
gender bias. Whether we talk about the political world, lifestyle or corporate
world, some generic differences are observed regarding the involvement of both
the groups. This differential distribution, being a part of the society at
large, exhibits its presence in the recorded data as well. Machine learning is
almost entirely dependent on the availability of data; and the idea of learning
from data and making predictions assumes that data defines the expected
behavior at large. Hence, with biased data the resulting models are corrupted
with those inherent biases too; and with the current popularity of ML in
products, this can result in a huge obstacle in the path of equality and
justice. This work studies and attempts to alleviate gender bias issues from
language vision models particularly the task of image captioning. We study the
extent of the impact of gender bias in existing datasets and propose a
methodology to mitigate its impact in caption based language vision models.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 04:33:44 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Aggarwal",
"Lavisha",
""
],
[
"Bhargava",
"Shruti",
""
]
] | TITLE: Fairness in AI Systems: Mitigating gender bias from language-vision
models
ABSTRACT: Our society is plagued by several biases, including racial biases, caste
biases, and gender bias. As a matter of fact, several years ago, most of these
notions were unheard of. These biases passed through generations along with
amplification have lead to scenarios where these have taken the role of
expected norms by certain groups in the society. One notable example is of
gender bias. Whether we talk about the political world, lifestyle or corporate
world, some generic differences are observed regarding the involvement of both
the groups. This differential distribution, being a part of the society at
large, exhibits its presence in the recorded data as well. Machine learning is
almost entirely dependent on the availability of data; and the idea of learning
from data and making predictions assumes that data defines the expected
behavior at large. Hence, with biased data the resulting models are corrupted
with those inherent biases too; and with the current popularity of ML in
products, this can result in a huge obstacle in the path of equality and
justice. This work studies and attempts to alleviate gender bias issues from
language vision models particularly the task of image captioning. We study the
extent of the impact of gender bias in existing datasets and propose a
methodology to mitigate its impact in caption based language vision models.
| no_new_dataset | 0.709069 |
2305.01910 | YuXuan Liu | YuXuan Liu, Nikhil Mishra, Pieter Abbeel, Xi Chen | Distributional Instance Segmentation: Modeling Uncertainty and High
Confidence Predictions with Latent-MaskRCNN | In ICRA 2023. Code and dataset are available at
https://segm.yuxuanliu.com/ | null | null | null | cs.CV cs.AI cs.RO | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Object recognition and instance segmentation are fundamental skills in any
robotic or autonomous system. Existing state-of-the-art methods are often
unable to capture meaningful uncertainty in challenging or ambiguous scenes,
and as such can cause critical errors in high-performance applications. In this
paper, we explore a class of distributional instance segmentation models using
latent codes that can model uncertainty over plausible hypotheses of object
masks. For robotic picking applications, we propose a confidence mask method to
achieve the high precision necessary in industrial use cases. We show that our
method can significantly reduce critical errors in robotic systems, including
our newly released dataset of ambiguous scenes in a robotic application. On a
real-world apparel-picking robot, our method significantly reduces double pick
errors while maintaining high performance.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 05:57:29 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Liu",
"YuXuan",
""
],
[
"Mishra",
"Nikhil",
""
],
[
"Abbeel",
"Pieter",
""
],
[
"Chen",
"Xi",
""
]
] | TITLE: Distributional Instance Segmentation: Modeling Uncertainty and High
Confidence Predictions with Latent-MaskRCNN
ABSTRACT: Object recognition and instance segmentation are fundamental skills in any
robotic or autonomous system. Existing state-of-the-art methods are often
unable to capture meaningful uncertainty in challenging or ambiguous scenes,
and as such can cause critical errors in high-performance applications. In this
paper, we explore a class of distributional instance segmentation models using
latent codes that can model uncertainty over plausible hypotheses of object
masks. For robotic picking applications, we propose a confidence mask method to
achieve the high precision necessary in industrial use cases. We show that our
method can significantly reduce critical errors in robotic systems, including
our newly released dataset of ambiguous scenes in a robotic application. On a
real-world apparel-picking robot, our method significantly reduces double pick
errors while maintaining high performance.
| new_dataset | 0.711262 |
2305.01911 | Lin Jiang | Lin Jiang, Anthony Dowling, Ming-C. Cheng, Yu Liu | PODTherm-GP: A Physics-based Data-Driven Approach for Effective
Architecture-Level Thermal Simulation of Multi-Core CPUs | null | null | null | null | cs.CE physics.comp-ph | http://creativecommons.org/licenses/by-nc-sa/4.0/ | A thermal simulation methodology derived from the proper orthogonal
decomposition (POD) and the Galerkin projection (GP), hereafter referred to as
PODTherm-GP, is evaluated in terms of its efficiency and accuracy in a
multi-core CPU. The GP projects the heat transfer equation onto a mathematical
space whose basis functions are generated from thermal data enabled by the POD
learning algorithm. The thermal solution data are collected from FEniCS using
the finite element method (FEM) accounting for appropriate parametric
variations. The GP incorporates physical principles of heat transfer in the
methodology to reach high accuracy and efficiency. The dynamic power map for
the CPU in FEM thermal simulation is generated from gem5 and McPACT, together
with the SPLASH-2 benchmarks as the simulation workload. It is shown that
PODTherm-GP offers an accurate thermal prediction of the CPU with a resolution
as fine as the FEM. It is also demonstrated that PODTherm-GP is capable of
predicting the dynamic thermal profile of the chip with a good accuracy beyond
the training conditions. Additionally, the approach offers a reduction in
degrees of freedom by more than 5 orders of magnitude and a speedup of 4
orders, compared to the FEM.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 05:59:23 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Jiang",
"Lin",
""
],
[
"Dowling",
"Anthony",
""
],
[
"Cheng",
"Ming-C.",
""
],
[
"Liu",
"Yu",
""
]
] | TITLE: PODTherm-GP: A Physics-based Data-Driven Approach for Effective
Architecture-Level Thermal Simulation of Multi-Core CPUs
ABSTRACT: A thermal simulation methodology derived from the proper orthogonal
decomposition (POD) and the Galerkin projection (GP), hereafter referred to as
PODTherm-GP, is evaluated in terms of its efficiency and accuracy in a
multi-core CPU. The GP projects the heat transfer equation onto a mathematical
space whose basis functions are generated from thermal data enabled by the POD
learning algorithm. The thermal solution data are collected from FEniCS using
the finite element method (FEM) accounting for appropriate parametric
variations. The GP incorporates physical principles of heat transfer in the
methodology to reach high accuracy and efficiency. The dynamic power map for
the CPU in FEM thermal simulation is generated from gem5 and McPACT, together
with the SPLASH-2 benchmarks as the simulation workload. It is shown that
PODTherm-GP offers an accurate thermal prediction of the CPU with a resolution
as fine as the FEM. It is also demonstrated that PODTherm-GP is capable of
predicting the dynamic thermal profile of the chip with a good accuracy beyond
the training conditions. Additionally, the approach offers a reduction in
degrees of freedom by more than 5 orders of magnitude and a speedup of 4
orders, compared to the FEM.
| no_new_dataset | 0.709617 |
2305.01914 | Yongbin Liu | Zhen Yang, Yongbin Liu, Chunping Ouyang | Causal Interventions-based Few-Shot Named Entity Recognition | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Few-shot named entity recognition (NER) systems aims at recognizing new
classes of entities based on a few labeled samples. A significant challenge in
the few-shot regime is prone to overfitting than the tasks with abundant
samples. The heavy overfitting in few-shot learning is mainly led by spurious
correlation caused by the few samples selection bias. To alleviate the problem
of the spurious correlation in the few-shot NER, in this paper, we propose a
causal intervention-based few-shot NER method. Based on the prototypical
network, the method intervenes in the context and prototype via backdoor
adjustment during training. In particular, intervening in the context of the
one-shot scenario is very difficult, so we intervene in the prototype via
incremental learning, which can also avoid catastrophic forgetting. Our
experiments on different benchmarks show that our approach achieves new
state-of-the-art results (achieving up to 29% absolute improvement and 12% on
average for all tasks).
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 06:11:39 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Yang",
"Zhen",
""
],
[
"Liu",
"Yongbin",
""
],
[
"Ouyang",
"Chunping",
""
]
] | TITLE: Causal Interventions-based Few-Shot Named Entity Recognition
ABSTRACT: Few-shot named entity recognition (NER) systems aims at recognizing new
classes of entities based on a few labeled samples. A significant challenge in
the few-shot regime is prone to overfitting than the tasks with abundant
samples. The heavy overfitting in few-shot learning is mainly led by spurious
correlation caused by the few samples selection bias. To alleviate the problem
of the spurious correlation in the few-shot NER, in this paper, we propose a
causal intervention-based few-shot NER method. Based on the prototypical
network, the method intervenes in the context and prototype via backdoor
adjustment during training. In particular, intervening in the context of the
one-shot scenario is very difficult, so we intervene in the prototype via
incremental learning, which can also avoid catastrophic forgetting. Our
experiments on different benchmarks show that our approach achieves new
state-of-the-art results (achieving up to 29% absolute improvement and 12% on
average for all tasks).
| no_new_dataset | 0.712445 |
2305.01936 | Georgios Batsis | Georgios Batsis, Ioannis Mademlis, Georgios Th. Papadopoulos | Illicit item detection in X-ray images for security applications | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Automated detection of contraband items in X-ray images can significantly
increase public safety, by enhancing the productivity and alleviating the
mental load of security officers in airports, subways, customs/post offices,
etc. The large volume and high throughput of passengers, mailed parcels, etc.,
during rush hours make it a Big Data analysis task. Modern computer vision
algorithms relying on Deep Neural Networks (DNNs) have proven capable of
undertaking this task even under resource-constrained and embedded execution
scenarios, e.g., as is the case with fast, single-stage, anchor-based object
detectors. This paper proposes a two-fold improvement of such algorithms for
the X-ray analysis domain, introducing two complementary novelties. Firstly,
more efficient anchors are obtained by hierarchical clustering the sizes of the
ground-truth training set bounding boxes; thus, the resulting anchors follow a
natural hierarchy aligned with the semantic structure of the data. Secondly,
the default Non-Maximum Suppression (NMS) algorithm at the end of the object
detection pipeline is modified to better handle occluded object detection and
to reduce the number of false predictions, by inserting the Efficient
Intersection over Union (E-IoU) metric into the Weighted Cluster NMS method.
E-IoU provides more discriminative geometrical correlations between the
candidate bounding boxes/Regions-of-Interest (RoIs). The proposed method is
implemented on a common single-stage object detector (YOLOv5) and its
experimental evaluation on a relevant public dataset indicates significant
accuracy gains over both the baseline and competing approaches. This highlights
the potential of Big Data analysis in enhancing public safety.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 07:28:05 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Batsis",
"Georgios",
""
],
[
"Mademlis",
"Ioannis",
""
],
[
"Papadopoulos",
"Georgios Th.",
""
]
] | TITLE: Illicit item detection in X-ray images for security applications
ABSTRACT: Automated detection of contraband items in X-ray images can significantly
increase public safety, by enhancing the productivity and alleviating the
mental load of security officers in airports, subways, customs/post offices,
etc. The large volume and high throughput of passengers, mailed parcels, etc.,
during rush hours make it a Big Data analysis task. Modern computer vision
algorithms relying on Deep Neural Networks (DNNs) have proven capable of
undertaking this task even under resource-constrained and embedded execution
scenarios, e.g., as is the case with fast, single-stage, anchor-based object
detectors. This paper proposes a two-fold improvement of such algorithms for
the X-ray analysis domain, introducing two complementary novelties. Firstly,
more efficient anchors are obtained by hierarchical clustering the sizes of the
ground-truth training set bounding boxes; thus, the resulting anchors follow a
natural hierarchy aligned with the semantic structure of the data. Secondly,
the default Non-Maximum Suppression (NMS) algorithm at the end of the object
detection pipeline is modified to better handle occluded object detection and
to reduce the number of false predictions, by inserting the Efficient
Intersection over Union (E-IoU) metric into the Weighted Cluster NMS method.
E-IoU provides more discriminative geometrical correlations between the
candidate bounding boxes/Regions-of-Interest (RoIs). The proposed method is
implemented on a common single-stage object detector (YOLOv5) and its
experimental evaluation on a relevant public dataset indicates significant
accuracy gains over both the baseline and competing approaches. This highlights
the potential of Big Data analysis in enhancing public safety.
| no_new_dataset | 0.707815 |
2305.01957 | Sardana Ivanova | Sardana Ivanova, Fredrik Aas Andreassen, Matias Jentoft, Sondre Wold,
Lilja {\O}vrelid | NorQuAD: Norwegian Question Answering Dataset | Accepted to NoDaLiDa 2023 | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In this paper we present NorQuAD: the first Norwegian question answering
dataset for machine reading comprehension. The dataset consists of 4,752
manually created question-answer pairs. We here detail the data collection
procedure and present statistics of the dataset. We also benchmark several
multilingual and Norwegian monolingual language models on the dataset and
compare them against human performance. The dataset will be made freely
available.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 08:17:07 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Ivanova",
"Sardana",
""
],
[
"Andreassen",
"Fredrik Aas",
""
],
[
"Jentoft",
"Matias",
""
],
[
"Wold",
"Sondre",
""
],
[
"Øvrelid",
"Lilja",
""
]
] | TITLE: NorQuAD: Norwegian Question Answering Dataset
ABSTRACT: In this paper we present NorQuAD: the first Norwegian question answering
dataset for machine reading comprehension. The dataset consists of 4,752
manually created question-answer pairs. We here detail the data collection
procedure and present statistics of the dataset. We also benchmark several
multilingual and Norwegian monolingual language models on the dataset and
compare them against human performance. The dataset will be made freely
available.
| new_dataset | 0.715989 |
2305.01965 | Mar\'ia Andrea Cruz Bland\'on | Mar\'ia Andrea Cruz Bland\'on, Alejandrina Cristia, Okko R\"as\"anen | Analysing the Impact of Audio Quality on the Use of Naturalistic
Long-Form Recordings for Infant-Directed Speech Research | Accepted at the CogSci2023 conference | null | null | null | cs.CL cs.SD eess.AS | http://creativecommons.org/licenses/by/4.0/ | Modelling of early language acquisition aims to understand how infants
bootstrap their language skills. The modelling encompasses properties of the
input data used for training the models, the cognitive hypotheses and their
algorithmic implementations being tested, and the evaluation methodologies to
compare models to human data. Recent developments have enabled the use of more
naturalistic training data for computational models. This also motivates
development of more naturalistic tests of model behaviour. A crucial step
towards such an aim is to develop representative speech datasets consisting of
speech heard by infants in their natural environments. However, a major
drawback of such recordings is that they are typically noisy, and it is
currently unclear how the sound quality could affect analyses and modelling
experiments conducted on such data. In this paper, we explore this aspect for
the case of infant-directed speech (IDS) and adult-directed speech (ADS)
analysis. First, we manually and automatically annotated audio quality of
utterances extracted from two corpora of child-centred long-form recordings (in
English and French). We then compared acoustic features of IDS and ADS in an
in-lab dataset and across different audio quality subsets of naturalistic data.
Finally, we assessed how the audio quality and recording environment may change
the conclusions of a modelling analysis using a recent self-supervised learning
model. Our results show that the use of modest and high audio quality
naturalistic speech data result in largely similar conclusions on IDS and ADS
in terms of acoustic analyses and modelling experiments. We also found that an
automatic sound quality assessment tool can be used to screen out useful parts
of long-form recordings for a closer analysis with comparable results to that
of manual quality annotation.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 08:25:37 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Blandón",
"María Andrea Cruz",
""
],
[
"Cristia",
"Alejandrina",
""
],
[
"Räsänen",
"Okko",
""
]
] | TITLE: Analysing the Impact of Audio Quality on the Use of Naturalistic
Long-Form Recordings for Infant-Directed Speech Research
ABSTRACT: Modelling of early language acquisition aims to understand how infants
bootstrap their language skills. The modelling encompasses properties of the
input data used for training the models, the cognitive hypotheses and their
algorithmic implementations being tested, and the evaluation methodologies to
compare models to human data. Recent developments have enabled the use of more
naturalistic training data for computational models. This also motivates
development of more naturalistic tests of model behaviour. A crucial step
towards such an aim is to develop representative speech datasets consisting of
speech heard by infants in their natural environments. However, a major
drawback of such recordings is that they are typically noisy, and it is
currently unclear how the sound quality could affect analyses and modelling
experiments conducted on such data. In this paper, we explore this aspect for
the case of infant-directed speech (IDS) and adult-directed speech (ADS)
analysis. First, we manually and automatically annotated audio quality of
utterances extracted from two corpora of child-centred long-form recordings (in
English and French). We then compared acoustic features of IDS and ADS in an
in-lab dataset and across different audio quality subsets of naturalistic data.
Finally, we assessed how the audio quality and recording environment may change
the conclusions of a modelling analysis using a recent self-supervised learning
model. Our results show that the use of modest and high audio quality
naturalistic speech data result in largely similar conclusions on IDS and ADS
in terms of acoustic analyses and modelling experiments. We also found that an
automatic sound quality assessment tool can be used to screen out useful parts
of long-form recordings for a closer analysis with comparable results to that
of manual quality annotation.
| no_new_dataset | 0.690533 |
2305.01968 | Min Cen | Min Cen, Xingyu Li, Bangwei Guo, Jitendra Jonnagaddala, Hong Zhang, Xu
Steven Xu | DPSeq: A Novel and Efficient Digital Pathology Classifier for Predicting
Cancer Biomarkers using Sequencer Architecture | null | null | null | null | eess.IV cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In digital pathology tasks, transformers have achieved state-of-the-art
results, surpassing convolutional neural networks (CNNs). However, transformers
are usually complex and resource intensive. In this study, we developed a novel
and efficient digital pathology classifier called DPSeq, to predict cancer
biomarkers through fine-tuning a sequencer architecture integrating horizon and
vertical bidirectional long short-term memory (BiLSTM) networks. Using
hematoxylin and eosin (H&E)-stained histopathological images of colorectal
cancer (CRC) from two international datasets: The Cancer Genome Atlas (TCGA)
and Molecular and Cellular Oncology (MCO), the predictive performance of DPSeq
was evaluated in series of experiments. DPSeq demonstrated exceptional
performance for predicting key biomarkers in CRC (MSI status, Hypermutation,
CIMP status, BRAF mutation, TP53 mutation and chromosomal instability [CING]),
outperforming most published state-of-the-art classifiers in a within-cohort
internal validation and a cross-cohort external validation. Additionally, under
the same experimental conditions using the same set of training and testing
datasets, DPSeq surpassed 4 CNN (ResNet18, ResNet50, MobileNetV2, and
EfficientNet) and 2 transformer (ViT and Swin-T) models, achieving the highest
AUROC and AUPRC values in predicting MSI status, BRAF mutation, and CIMP
status. Furthermore, DPSeq required less time for both training and prediction
due to its simple architecture. Therefore, DPSeq appears to be the preferred
choice over transformer and CNN models for predicting cancer biomarkers.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 08:31:44 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Cen",
"Min",
""
],
[
"Li",
"Xingyu",
""
],
[
"Guo",
"Bangwei",
""
],
[
"Jonnagaddala",
"Jitendra",
""
],
[
"Zhang",
"Hong",
""
],
[
"Xu",
"Xu Steven",
""
]
] | TITLE: DPSeq: A Novel and Efficient Digital Pathology Classifier for Predicting
Cancer Biomarkers using Sequencer Architecture
ABSTRACT: In digital pathology tasks, transformers have achieved state-of-the-art
results, surpassing convolutional neural networks (CNNs). However, transformers
are usually complex and resource intensive. In this study, we developed a novel
and efficient digital pathology classifier called DPSeq, to predict cancer
biomarkers through fine-tuning a sequencer architecture integrating horizon and
vertical bidirectional long short-term memory (BiLSTM) networks. Using
hematoxylin and eosin (H&E)-stained histopathological images of colorectal
cancer (CRC) from two international datasets: The Cancer Genome Atlas (TCGA)
and Molecular and Cellular Oncology (MCO), the predictive performance of DPSeq
was evaluated in series of experiments. DPSeq demonstrated exceptional
performance for predicting key biomarkers in CRC (MSI status, Hypermutation,
CIMP status, BRAF mutation, TP53 mutation and chromosomal instability [CING]),
outperforming most published state-of-the-art classifiers in a within-cohort
internal validation and a cross-cohort external validation. Additionally, under
the same experimental conditions using the same set of training and testing
datasets, DPSeq surpassed 4 CNN (ResNet18, ResNet50, MobileNetV2, and
EfficientNet) and 2 transformer (ViT and Swin-T) models, achieving the highest
AUROC and AUPRC values in predicting MSI status, BRAF mutation, and CIMP
status. Furthermore, DPSeq required less time for both training and prediction
due to its simple architecture. Therefore, DPSeq appears to be the preferred
choice over transformer and CNN models for predicting cancer biomarkers.
| no_new_dataset | 0.712445 |
2305.02032 | Sajid Javed Dr | Sajid Javed, Arif Mahmood, Talha Qaiser, Naoufel Werghi, and Nasir
Rajpoot | Unsupervised Mutual Transformer Learning for Multi-Gigapixel Whole Slide
Image Classification | null | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Classification of gigapixel Whole Slide Images (WSIs) is an important
prediction task in the emerging area of computational pathology. There has been
a surge of research in deep learning models for WSI classification with
clinical applications such as cancer detection or prediction of molecular
mutations from WSIs. Most methods require expensive and labor-intensive manual
annotations by expert pathologists. Weakly supervised Multiple Instance
Learning (MIL) methods have recently demonstrated excellent performance;
however, they still require large slide-level labeled training datasets that
need a careful inspection of each slide by an expert pathologist. In this work,
we propose a fully unsupervised WSI classification algorithm based on mutual
transformer learning. Instances from gigapixel WSI (i.e., image patches) are
transformed into a latent space and then inverse-transformed to the original
space. Using the transformation loss, pseudo-labels are generated and cleaned
using a transformer label-cleaner. The proposed transformer-based pseudo-label
generation and cleaning modules mutually train each other iteratively in an
unsupervised manner. A discriminative learning mechanism is introduced to
improve normal versus cancerous instance labeling. In addition to unsupervised
classification, we demonstrate the effectiveness of the proposed framework for
weak supervision for cancer subtype classification as downstream analysis.
Extensive experiments on four publicly available datasets show excellent
performance compared to the state-of-the-art methods. We intend to make the
source code of our algorithm publicly available soon.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 10:54:18 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Javed",
"Sajid",
""
],
[
"Mahmood",
"Arif",
""
],
[
"Qaiser",
"Talha",
""
],
[
"Werghi",
"Naoufel",
""
],
[
"Rajpoot",
"Nasir",
""
]
] | TITLE: Unsupervised Mutual Transformer Learning for Multi-Gigapixel Whole Slide
Image Classification
ABSTRACT: Classification of gigapixel Whole Slide Images (WSIs) is an important
prediction task in the emerging area of computational pathology. There has been
a surge of research in deep learning models for WSI classification with
clinical applications such as cancer detection or prediction of molecular
mutations from WSIs. Most methods require expensive and labor-intensive manual
annotations by expert pathologists. Weakly supervised Multiple Instance
Learning (MIL) methods have recently demonstrated excellent performance;
however, they still require large slide-level labeled training datasets that
need a careful inspection of each slide by an expert pathologist. In this work,
we propose a fully unsupervised WSI classification algorithm based on mutual
transformer learning. Instances from gigapixel WSI (i.e., image patches) are
transformed into a latent space and then inverse-transformed to the original
space. Using the transformation loss, pseudo-labels are generated and cleaned
using a transformer label-cleaner. The proposed transformer-based pseudo-label
generation and cleaning modules mutually train each other iteratively in an
unsupervised manner. A discriminative learning mechanism is introduced to
improve normal versus cancerous instance labeling. In addition to unsupervised
classification, we demonstrate the effectiveness of the proposed framework for
weak supervision for cancer subtype classification as downstream analysis.
Extensive experiments on four publicly available datasets show excellent
performance compared to the state-of-the-art methods. We intend to make the
source code of our algorithm publicly available soon.
| no_new_dataset | 0.709598 |
2305.02052 | Marc Carrascosa-Zamacois | Marc Carrascosa, Giovanni Geraci, Edward Knightly, Boris Bellalta | Wi-Fi Multi-Link Operation: An Experimental Study of Latency and
Throughput | null | null | null | null | cs.NI cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this article, we investigate the real-world capability of the multi-link
operation (MLO) framework -- one of the key MAC-layer features included in the
IEEE 802.11be amendment -- by using a large dataset containing 5 GHz spectrum
occupancy measurements on multiple channels. Our results show that when both
available links are often busy, as is the case in ultra-dense and crowded
scenarios, MLO attains the highest throughput gains over single-link operation
(SLO) since it is able to leverage multiple intermittent transmission
opportunities. As for latency, if the two links exhibit statistically the same
level of occupancy, MLO can outperform SLO by one order of magnitude. In
contrast, in asymmetrically occupied links, MLO can sometimes be detrimental
and even increase latency. We study this somewhat unexpected phenomenon, and
find its origins to be packets suboptimally mapped to either link before
carrying out the backoff, with the latter likely to be interrupted on the
busier link. We cross validate our study with real-time traffic generated by a
cloud gaming application and quantify MLO's benefits for latency-sensitive
applications.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 11:38:34 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Carrascosa",
"Marc",
""
],
[
"Geraci",
"Giovanni",
""
],
[
"Knightly",
"Edward",
""
],
[
"Bellalta",
"Boris",
""
]
] | TITLE: Wi-Fi Multi-Link Operation: An Experimental Study of Latency and
Throughput
ABSTRACT: In this article, we investigate the real-world capability of the multi-link
operation (MLO) framework -- one of the key MAC-layer features included in the
IEEE 802.11be amendment -- by using a large dataset containing 5 GHz spectrum
occupancy measurements on multiple channels. Our results show that when both
available links are often busy, as is the case in ultra-dense and crowded
scenarios, MLO attains the highest throughput gains over single-link operation
(SLO) since it is able to leverage multiple intermittent transmission
opportunities. As for latency, if the two links exhibit statistically the same
level of occupancy, MLO can outperform SLO by one order of magnitude. In
contrast, in asymmetrically occupied links, MLO can sometimes be detrimental
and even increase latency. We study this somewhat unexpected phenomenon, and
find its origins to be packets suboptimally mapped to either link before
carrying out the backoff, with the latter likely to be interrupted on the
busier link. We cross validate our study with real-time traffic generated by a
cloud gaming application and quantify MLO's benefits for latency-sensitive
applications.
| no_new_dataset | 0.70878 |
2305.02061 | Ahmed N. Ahmed | Ahmed N. Ahmed, Siegfried Mercelis, Ali Anwar | Attention Based Feature Fusion For Multi-Agent Collaborative Perception | null | null | null | null | cs.MA cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the domain of intelligent transportation systems (ITS), collaborative
perception has emerged as a promising approach to overcome the limitations of
individual perception by enabling multiple agents to exchange information, thus
enhancing their situational awareness. Collaborative perception overcomes the
limitations of individual sensors, allowing connected agents to perceive
environments beyond their line-of-sight and field of view. However, the
reliability of collaborative perception heavily depends on the data aggregation
strategy and communication bandwidth, which must overcome the challenges posed
by limited network resources. To improve the precision of object detection and
alleviate limited network resources, we propose an intermediate collaborative
perception solution in the form of a graph attention network (GAT). The
proposed approach develops an attention-based aggregation strategy to fuse
intermediate representations exchanged among multiple connected agents. This
approach adaptively highlights important regions in the intermediate feature
maps at both the channel and spatial levels, resulting in improved object
detection precision. We propose a feature fusion scheme using attention-based
architectures and evaluate the results quantitatively in comparison to other
state-of-the-art collaborative perception approaches. Our proposed approach is
validated using the V2XSim dataset. The results of this work demonstrate the
efficacy of the proposed approach for intermediate collaborative perception in
improving object detection average precision while reducing network resource
usage.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 12:06:11 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Ahmed",
"Ahmed N.",
""
],
[
"Mercelis",
"Siegfried",
""
],
[
"Anwar",
"Ali",
""
]
] | TITLE: Attention Based Feature Fusion For Multi-Agent Collaborative Perception
ABSTRACT: In the domain of intelligent transportation systems (ITS), collaborative
perception has emerged as a promising approach to overcome the limitations of
individual perception by enabling multiple agents to exchange information, thus
enhancing their situational awareness. Collaborative perception overcomes the
limitations of individual sensors, allowing connected agents to perceive
environments beyond their line-of-sight and field of view. However, the
reliability of collaborative perception heavily depends on the data aggregation
strategy and communication bandwidth, which must overcome the challenges posed
by limited network resources. To improve the precision of object detection and
alleviate limited network resources, we propose an intermediate collaborative
perception solution in the form of a graph attention network (GAT). The
proposed approach develops an attention-based aggregation strategy to fuse
intermediate representations exchanged among multiple connected agents. This
approach adaptively highlights important regions in the intermediate feature
maps at both the channel and spatial levels, resulting in improved object
detection precision. We propose a feature fusion scheme using attention-based
architectures and evaluate the results quantitatively in comparison to other
state-of-the-art collaborative perception approaches. Our proposed approach is
validated using the V2XSim dataset. The results of this work demonstrate the
efficacy of the proposed approach for intermediate collaborative perception in
improving object detection average precision while reducing network resource
usage.
| no_new_dataset | 0.710208 |
2305.02081 | Orestis Papakyriakopoulos | Orestis Papakyriakopoulos, Alice Xiang | Considerations for Ethical Speech Recognition Datasets | null | In Proceedings of the Sixteenth ACM International Conference on
Web Search and Data Mining (WSDM '23), February 27-March 3, 2023, Singapore,
Singapore. ACM, New York, NY, USA, 2 pages.
https://doi.org/10.1145/3539597.3575793 | 10.1145/3539597.3575793 | null | cs.CY cs.CL cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Speech AI Technologies are largely trained on publicly available datasets or
by the massive web-crawling of speech. In both cases, data acquisition focuses
on minimizing collection effort, without necessarily taking the data subjects'
protection or user needs into consideration. This results to models that are
not robust when used on users who deviate from the dominant demographics in the
training set, discriminating individuals having different dialects, accents,
speaking styles, and disfluencies. In this talk, we use automatic speech
recognition as a case study and examine the properties that ethical speech
datasets should possess towards responsible AI applications. We showcase
diversity issues, inclusion practices, and necessary considerations that can
improve trained models, while facilitating model explainability and protecting
users and data subjects. We argue for the legal & privacy protection of data
subjects, targeted data sampling corresponding to user demographics & needs,
appropriate meta data that ensure explainability & accountability in cases of
model failure, and the sociotechnical \& situated model design. We hope this
talk can inspire researchers \& practitioners to design and use more
human-centric datasets in speech technologies and other domains, in ways that
empower and respect users, while improving machine learning models' robustness
and utility.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 12:38:14 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Papakyriakopoulos",
"Orestis",
""
],
[
"Xiang",
"Alice",
""
]
] | TITLE: Considerations for Ethical Speech Recognition Datasets
ABSTRACT: Speech AI Technologies are largely trained on publicly available datasets or
by the massive web-crawling of speech. In both cases, data acquisition focuses
on minimizing collection effort, without necessarily taking the data subjects'
protection or user needs into consideration. This results to models that are
not robust when used on users who deviate from the dominant demographics in the
training set, discriminating individuals having different dialects, accents,
speaking styles, and disfluencies. In this talk, we use automatic speech
recognition as a case study and examine the properties that ethical speech
datasets should possess towards responsible AI applications. We showcase
diversity issues, inclusion practices, and necessary considerations that can
improve trained models, while facilitating model explainability and protecting
users and data subjects. We argue for the legal & privacy protection of data
subjects, targeted data sampling corresponding to user demographics & needs,
appropriate meta data that ensure explainability & accountability in cases of
model failure, and the sociotechnical \& situated model design. We hope this
talk can inspire researchers \& practitioners to design and use more
human-centric datasets in speech technologies and other domains, in ways that
empower and respect users, while improving machine learning models' robustness
and utility.
| no_new_dataset | 0.703982 |
2305.02085 | Maloy Kumar Devnath | Maloy Kumar Devnath, Avijoy Chakma, Mohammad Saeid Anwar, Emon Dey,
Zahid Hasan, Marc Conn, Biplab Pal, Nirmalya Roy | A Systematic Study on Object Recognition Using Millimeter-wave Radar | null | null | null | null | cs.CV cs.AI cs.CE cs.RO | http://creativecommons.org/licenses/by/4.0/ | Due to its light and weather-independent sensing, millimeter-wave (MMW) radar
is essential in smart environments. Intelligent vehicle systems and
industry-grade MMW radars have integrated such capabilities. Industry-grade MMW
radars are expensive and hard to get for community-purpose smart environment
applications. However, commercially available MMW radars have hidden
underpinning challenges that need to be investigated for tasks like recognizing
objects and activities, real-time person tracking, object localization, etc.
Image and video data are straightforward to gather, understand, and annotate
for such jobs. Image and video data are light and weather-dependent,
susceptible to the occlusion effect, and present privacy problems. To eliminate
dependence and ensure privacy, commercial MMW radars should be tested. MMW
radar's practicality and performance in varied operating settings must be
addressed before promoting it. To address the problems, we collected a dataset
using Texas Instruments' Automotive mmWave Radar (AWR2944) and reported the
best experimental settings for object recognition performance using different
deep learning algorithms. Our extensive data gathering technique allows us to
systematically explore and identify object identification task problems under
cross-ambience conditions. We investigated several solutions and published
detailed experimental data.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 12:42:44 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Devnath",
"Maloy Kumar",
""
],
[
"Chakma",
"Avijoy",
""
],
[
"Anwar",
"Mohammad Saeid",
""
],
[
"Dey",
"Emon",
""
],
[
"Hasan",
"Zahid",
""
],
[
"Conn",
"Marc",
""
],
[
"Pal",
"Biplab",
""
],
[
"Roy",
"Nirmalya",
""
]
] | TITLE: A Systematic Study on Object Recognition Using Millimeter-wave Radar
ABSTRACT: Due to its light and weather-independent sensing, millimeter-wave (MMW) radar
is essential in smart environments. Intelligent vehicle systems and
industry-grade MMW radars have integrated such capabilities. Industry-grade MMW
radars are expensive and hard to get for community-purpose smart environment
applications. However, commercially available MMW radars have hidden
underpinning challenges that need to be investigated for tasks like recognizing
objects and activities, real-time person tracking, object localization, etc.
Image and video data are straightforward to gather, understand, and annotate
for such jobs. Image and video data are light and weather-dependent,
susceptible to the occlusion effect, and present privacy problems. To eliminate
dependence and ensure privacy, commercial MMW radars should be tested. MMW
radar's practicality and performance in varied operating settings must be
addressed before promoting it. To address the problems, we collected a dataset
using Texas Instruments' Automotive mmWave Radar (AWR2944) and reported the
best experimental settings for object recognition performance using different
deep learning algorithms. Our extensive data gathering technique allows us to
systematically explore and identify object identification task problems under
cross-ambience conditions. We investigated several solutions and published
detailed experimental data.
| new_dataset | 0.705011 |
2305.02103 | Andrea Ramazzina | Andrea Ramazzina, Mario Bijelic, Stefanie Walz, Alessandro Sanvito,
Dominik Scheuble and Felix Heide | ScatterNeRF: Seeing Through Fog with Physically-Based Inverse Neural
Rendering | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vision in adverse weather conditions, whether it be snow, rain, or fog is
challenging. In these scenarios, scattering and attenuation severly degrades
image quality. Handling such inclement weather conditions, however, is
essential to operate autonomous vehicles, drones and robotic applications where
human performance is impeded the most. A large body of work explores removing
weather-induced image degradations with dehazing methods. Most methods rely on
single images as input and struggle to generalize from synthetic
fully-supervised training approaches or to generate high fidelity results from
unpaired real-world datasets. With data as bottleneck and most of today's
training data relying on good weather conditions with inclement weather as
outlier, we rely on an inverse rendering approach to reconstruct the scene
content. We introduce ScatterNeRF, a neural rendering method which adequately
renders foggy scenes and decomposes the fog-free background from the
participating media-exploiting the multiple views from a short automotive
sequence without the need for a large training data corpus. Instead, the
rendering approach is optimized on the multi-view scene itself, which can be
typically captured by an autonomous vehicle, robot or drone during operation.
Specifically, we propose a disentangled representation for the scattering
volume and the scene objects, and learn the scene reconstruction with
physics-inspired losses. We validate our method by capturing multi-view
In-the-Wild data and controlled captures in a large-scale fog chamber.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 13:24:06 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Ramazzina",
"Andrea",
""
],
[
"Bijelic",
"Mario",
""
],
[
"Walz",
"Stefanie",
""
],
[
"Sanvito",
"Alessandro",
""
],
[
"Scheuble",
"Dominik",
""
],
[
"Heide",
"Felix",
""
]
] | TITLE: ScatterNeRF: Seeing Through Fog with Physically-Based Inverse Neural
Rendering
ABSTRACT: Vision in adverse weather conditions, whether it be snow, rain, or fog is
challenging. In these scenarios, scattering and attenuation severly degrades
image quality. Handling such inclement weather conditions, however, is
essential to operate autonomous vehicles, drones and robotic applications where
human performance is impeded the most. A large body of work explores removing
weather-induced image degradations with dehazing methods. Most methods rely on
single images as input and struggle to generalize from synthetic
fully-supervised training approaches or to generate high fidelity results from
unpaired real-world datasets. With data as bottleneck and most of today's
training data relying on good weather conditions with inclement weather as
outlier, we rely on an inverse rendering approach to reconstruct the scene
content. We introduce ScatterNeRF, a neural rendering method which adequately
renders foggy scenes and decomposes the fog-free background from the
participating media-exploiting the multiple views from a short automotive
sequence without the need for a large training data corpus. Instead, the
rendering approach is optimized on the multi-view scene itself, which can be
typically captured by an autonomous vehicle, robot or drone during operation.
Specifically, we propose a disentangled representation for the scattering
volume and the scene objects, and learn the scene reconstruction with
physics-inspired losses. We validate our method by capturing multi-view
In-the-Wild data and controlled captures in a large-scale fog chamber.
| no_new_dataset | 0.711813 |
2305.02104 | Domenic Rosati | Domenic Rosati | Background Knowledge Grounding for Readable, Relevant, and Factual
Biomedical Lay Summaries | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Communication of scientific findings to the public is important for keeping
non-experts informed of developments such as life-saving medical treatments.
However, generating readable lay summaries from scientific documents is
challenging, and currently, these summaries suffer from critical factual
errors. One popular intervention for improving factuality is using additional
external knowledge to provide factual grounding. However, it is unclear how
these grounding sources should be retrieved, selected, or integrated, and how
supplementary grounding documents might affect the readability or relevance of
the generated summaries. We develop a simple method for selecting grounding
sources and integrating them with source documents. We then use the BioLaySum
summarization dataset to evaluate the effects of different grounding sources on
summary quality. We found that grounding source documents improves the
relevance and readability of lay summaries but does not improve factuality of
lay summaries. This continues to be true in zero-shot summarization settings
where we hypothesized that grounding might be even more important for factual
lay summaries.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 13:24:49 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Rosati",
"Domenic",
""
]
] | TITLE: Background Knowledge Grounding for Readable, Relevant, and Factual
Biomedical Lay Summaries
ABSTRACT: Communication of scientific findings to the public is important for keeping
non-experts informed of developments such as life-saving medical treatments.
However, generating readable lay summaries from scientific documents is
challenging, and currently, these summaries suffer from critical factual
errors. One popular intervention for improving factuality is using additional
external knowledge to provide factual grounding. However, it is unclear how
these grounding sources should be retrieved, selected, or integrated, and how
supplementary grounding documents might affect the readability or relevance of
the generated summaries. We develop a simple method for selecting grounding
sources and integrating them with source documents. We then use the BioLaySum
summarization dataset to evaluate the effects of different grounding sources on
summary quality. We found that grounding source documents improves the
relevance and readability of lay summaries but does not improve factuality of
lay summaries. This continues to be true in zero-shot summarization settings
where we hypothesized that grounding might be even more important for factual
lay summaries.
| no_new_dataset | 0.650717 |
2305.02116 | Zhen Wei | Zhen Wei and Pascal Fua and Micha\"el Bauerheim | Automatic Parameterization for Aerodynamic Shape Optimization via Deep
Geometric Learning | 15 pages, to be appeared at AIAA Aviation Forum 2023 | null | null | null | cs.CV physics.flu-dyn | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We propose two deep learning models that fully automate shape
parameterization for aerodynamic shape optimization. Both models are optimized
to parameterize via deep geometric learning to embed human prior knowledge into
learned geometric patterns, eliminating the need for further handcrafting. The
Latent Space Model (LSM) learns a low-dimensional latent representation of an
object from a dataset of various geometries, while the Direct Mapping Model
(DMM) builds parameterization on the fly using only one geometry of interest.
We also devise a novel regularization loss that efficiently integrates
volumetric mesh deformation into the parameterization model. The models
directly manipulate the high-dimensional mesh data by moving vertices. LSM and
DMM are fully differentiable, enabling gradient-based, end-to-end pipeline
design and plug-and-play deployment of surrogate models or adjoint solvers. We
perform shape optimization experiments on 2D airfoils and discuss the
applicable scenarios for the two models.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 13:45:40 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Wei",
"Zhen",
""
],
[
"Fua",
"Pascal",
""
],
[
"Bauerheim",
"Michaël",
""
]
] | TITLE: Automatic Parameterization for Aerodynamic Shape Optimization via Deep
Geometric Learning
ABSTRACT: We propose two deep learning models that fully automate shape
parameterization for aerodynamic shape optimization. Both models are optimized
to parameterize via deep geometric learning to embed human prior knowledge into
learned geometric patterns, eliminating the need for further handcrafting. The
Latent Space Model (LSM) learns a low-dimensional latent representation of an
object from a dataset of various geometries, while the Direct Mapping Model
(DMM) builds parameterization on the fly using only one geometry of interest.
We also devise a novel regularization loss that efficiently integrates
volumetric mesh deformation into the parameterization model. The models
directly manipulate the high-dimensional mesh data by moving vertices. LSM and
DMM are fully differentiable, enabling gradient-based, end-to-end pipeline
design and plug-and-play deployment of surrogate models or adjoint solvers. We
perform shape optimization experiments on 2D airfoils and discuss the
applicable scenarios for the two models.
| no_new_dataset | 0.710867 |
2305.02126 | Bahri Batuhan Bilecen | Bahri Batuhan Bilecen, Mustafa Ayazoglu | Bicubic++: Slim, Slimmer, Slimmest -- Designing an Industry-Grade
Super-Resolution Network | Winner of the New Trends in Image Restoration and Enhancement (NTIRE)
@ CVPR 2023, Real Time Super Resolution (RTSR) Challange Track 2 (x3
super-resolution). Code available at:
https://github.com/aselsan-research-imaging-team/bicubic-plusplus | null | null | null | cs.CV cs.LG eess.IV | http://creativecommons.org/licenses/by/4.0/ | We propose a real-time and lightweight single-image super-resolution (SR)
network named Bicubic++. Despite using spatial dimensions of the input image
across the whole network, Bicubic++ first learns quick reversible downgraded
and lower resolution features of the image in order to decrease the number of
computations. We also construct a training pipeline, where we apply an
end-to-end global structured pruning of convolutional layers without using
metrics like magnitude and gradient norms, and focus on optimizing the pruned
network's PSNR on the validation set. Furthermore, we have experimentally shown
that the bias terms take considerable amount of the runtime while increasing
PSNR marginally, hence we have also applied bias removal to the convolutional
layers. Our method adds ~1dB on Bicubic upscaling PSNR for all tested SR
datasets and runs with ~1.17ms on RTX3090 and ~2.9ms on RTX3070, for 720p
inputs and 4K outputs, both in FP16 precision. Bicubic++ won NTIRE 2023 RTSR
Track 2 x3 SR competition and is the fastest among all competitive methods.
Being almost as fast as the standard Bicubic upsampling method, we believe that
Bicubic++ can set a new industry standard.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 13:57:00 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Bilecen",
"Bahri Batuhan",
""
],
[
"Ayazoglu",
"Mustafa",
""
]
] | TITLE: Bicubic++: Slim, Slimmer, Slimmest -- Designing an Industry-Grade
Super-Resolution Network
ABSTRACT: We propose a real-time and lightweight single-image super-resolution (SR)
network named Bicubic++. Despite using spatial dimensions of the input image
across the whole network, Bicubic++ first learns quick reversible downgraded
and lower resolution features of the image in order to decrease the number of
computations. We also construct a training pipeline, where we apply an
end-to-end global structured pruning of convolutional layers without using
metrics like magnitude and gradient norms, and focus on optimizing the pruned
network's PSNR on the validation set. Furthermore, we have experimentally shown
that the bias terms take considerable amount of the runtime while increasing
PSNR marginally, hence we have also applied bias removal to the convolutional
layers. Our method adds ~1dB on Bicubic upscaling PSNR for all tested SR
datasets and runs with ~1.17ms on RTX3090 and ~2.9ms on RTX3070, for 720p
inputs and 4K outputs, both in FP16 precision. Bicubic++ won NTIRE 2023 RTSR
Track 2 x3 SR competition and is the fastest among all competitive methods.
Being almost as fast as the standard Bicubic upsampling method, we believe that
Bicubic++ can set a new industry standard.
| no_new_dataset | 0.710622 |
2305.02142 | Nabajeet Barman | Nabajeet Barman, Yuriy Reznik and Maria Martini | Datasheet for Subjective and Objective Quality Assessment Datasets | null | null | null | null | cs.MM | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Over the years, many subjective and objective quality assessment datasets
have been created and made available to the research community. However, there
is no standard process for documenting the various aspects of the dataset, such
as details about the source sequences, number of test subjects, test
methodology, encoding settings, etc. Such information is often of great
importance to the users of the dataset as it can help them get a quick
understanding of the motivation and scope of the dataset. Without such a
template, it is left to each reader to collate the information from the
relevant publication or website, which is a tedious and time-consuming process.
In some cases, the absence of a template to guide the documentation process can
result in an unintentional omission of some important information.
This paper addresses this simple but significant gap by proposing a datasheet
template for documenting various aspects of subjective and objective quality
assessment datasets for multimedia data. The contributions presented in this
work aim to simplify the documentation process for existing and new datasets
and improve their reproducibility. The proposed datasheet template is available
on GitHub, along with a few sample datasheets of a few open-source audiovisual
subjective and objective datasets.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 14:22:42 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Barman",
"Nabajeet",
""
],
[
"Reznik",
"Yuriy",
""
],
[
"Martini",
"Maria",
""
]
] | TITLE: Datasheet for Subjective and Objective Quality Assessment Datasets
ABSTRACT: Over the years, many subjective and objective quality assessment datasets
have been created and made available to the research community. However, there
is no standard process for documenting the various aspects of the dataset, such
as details about the source sequences, number of test subjects, test
methodology, encoding settings, etc. Such information is often of great
importance to the users of the dataset as it can help them get a quick
understanding of the motivation and scope of the dataset. Without such a
template, it is left to each reader to collate the information from the
relevant publication or website, which is a tedious and time-consuming process.
In some cases, the absence of a template to guide the documentation process can
result in an unintentional omission of some important information.
This paper addresses this simple but significant gap by proposing a datasheet
template for documenting various aspects of subjective and objective quality
assessment datasets for multimedia data. The contributions presented in this
work aim to simplify the documentation process for existing and new datasets
and improve their reproducibility. The proposed datasheet template is available
on GitHub, along with a few sample datasheets of a few open-source audiovisual
subjective and objective datasets.
| no_new_dataset | 0.706431 |
2305.02156 | Jimmy Lin | Xueguang Ma, Xinyu Zhang, Ronak Pradeep, Jimmy Lin | Zero-Shot Listwise Document Reranking with a Large Language Model | null | null | null | null | cs.IR cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Supervised ranking methods based on bi-encoder or cross-encoder architectures
have shown success in multi-stage text ranking tasks, but they require large
amounts of relevance judgments as training data. In this work, we propose
Listwise Reranker with a Large Language Model (LRL), which achieves strong
reranking effectiveness without using any task-specific training data.
Different from the existing pointwise ranking methods, where documents are
scored independently and ranked according to the scores, LRL directly generates
a reordered list of document identifiers given the candidate documents.
Experiments on three TREC web search datasets demonstrate that LRL not only
outperforms zero-shot pointwise methods when reranking first-stage retrieval
results, but can also act as a final-stage reranker to improve the top-ranked
results of a pointwise method for improved efficiency. Additionally, we apply
our approach to subsets of MIRACL, a recent multilingual retrieval dataset,
with results showing its potential to generalize across different languages.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 14:45:34 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Ma",
"Xueguang",
""
],
[
"Zhang",
"Xinyu",
""
],
[
"Pradeep",
"Ronak",
""
],
[
"Lin",
"Jimmy",
""
]
] | TITLE: Zero-Shot Listwise Document Reranking with a Large Language Model
ABSTRACT: Supervised ranking methods based on bi-encoder or cross-encoder architectures
have shown success in multi-stage text ranking tasks, but they require large
amounts of relevance judgments as training data. In this work, we propose
Listwise Reranker with a Large Language Model (LRL), which achieves strong
reranking effectiveness without using any task-specific training data.
Different from the existing pointwise ranking methods, where documents are
scored independently and ranked according to the scores, LRL directly generates
a reordered list of document identifiers given the candidate documents.
Experiments on three TREC web search datasets demonstrate that LRL not only
outperforms zero-shot pointwise methods when reranking first-stage retrieval
results, but can also act as a final-stage reranker to improve the top-ranked
results of a pointwise method for improved efficiency. Additionally, we apply
our approach to subsets of MIRACL, a recent multilingual retrieval dataset,
with results showing its potential to generalize across different languages.
| no_new_dataset | 0.709849 |
2305.02204 | Khanh Duy Nguyen | Khanh Duy Nguyen, Nima Shahbazi and Abolfazl Asudeh | PopSim: An Individual-level Population Simulator for Equitable
Allocation of City Resources | Published as part of the Workshop on Algorithmic Fairness in
Artificial Intelligence, Machine Learning, and Decision Making (AFair-AMLD)
at the SIAM International Conference on Data Mining (SDM23) | AFair-AMLD workshop at SIAM International Conference on Data
Mining (2023) | null | null | cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Historical systematic exclusionary tactics based on race have forced people
of certain demographic groups to congregate in specific urban areas. Aside from
the ethical aspects of such segregation, these policies have implications for
the allocation of urban resources including public transportation, healthcare,
and education within the cities. The initial step towards addressing these
issues involves conducting an audit to assess the status of equitable resource
allocation. However, due to privacy and confidentiality concerns,
individual-level data containing demographic information cannot be made
publicly available. By leveraging publicly available aggregated demographic
statistics data, we introduce PopSim, a system for generating semi-synthetic
individual-level population data with demographic information. We use PopSim to
generate multiple benchmark datasets for the city of Chicago and conduct
extensive statistical evaluations to validate those. We further use our
datasets for several case studies that showcase the application of our system
for auditing equitable allocation of city resources.
| [
{
"version": "v1",
"created": "Tue, 25 Apr 2023 23:43:21 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Nguyen",
"Khanh Duy",
""
],
[
"Shahbazi",
"Nima",
""
],
[
"Asudeh",
"Abolfazl",
""
]
] | TITLE: PopSim: An Individual-level Population Simulator for Equitable
Allocation of City Resources
ABSTRACT: Historical systematic exclusionary tactics based on race have forced people
of certain demographic groups to congregate in specific urban areas. Aside from
the ethical aspects of such segregation, these policies have implications for
the allocation of urban resources including public transportation, healthcare,
and education within the cities. The initial step towards addressing these
issues involves conducting an audit to assess the status of equitable resource
allocation. However, due to privacy and confidentiality concerns,
individual-level data containing demographic information cannot be made
publicly available. By leveraging publicly available aggregated demographic
statistics data, we introduce PopSim, a system for generating semi-synthetic
individual-level population data with demographic information. We use PopSim to
generate multiple benchmark datasets for the city of Chicago and conduct
extensive statistical evaluations to validate those. We further use our
datasets for several case studies that showcase the application of our system
for auditing equitable allocation of city resources.
| no_new_dataset | 0.550124 |
2305.02214 | Kai Xiong | Kai Xiong and Zhihong Wang and Supeng Leng and Jianhua He | A Digital Twin Empowered Lightweight Model Sharing Scheme for
Multi-Robot Systems | 16 pages, 12 figures, journal | null | null | null | cs.IT cs.RO cs.SY eess.SY math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-robot system for manufacturing is an Industry Internet of Things (IIoT)
paradigm with significant operational cost savings and productivity
improvement, where Unmanned Aerial Vehicles (UAVs) are employed to control and
implement collaborative productions without human intervention. This
mission-critical system relies on 3-Dimension (3-D) scene recognition to
improve operation accuracy in the production line and autonomous piloting.
However, implementing 3-D point cloud learning, such as Pointnet, is
challenging due to limited sensing and computing resources equipped with UAVs.
Therefore, we propose a Digital Twin (DT) empowered Knowledge Distillation (KD)
method to generate several lightweight learning models and select the optimal
model to deploy on UAVs. With a digital replica of the UAVs preserved at the
edge server, the DT system controls the model sharing network topology and
learning model structure to improve recognition accuracy further. Moreover, we
employ network calculus to formulate and solve the model sharing configuration
problem toward minimal resource consumption, as well as convergence. Simulation
experiments are conducted over a popular point cloud dataset to evaluate the
proposed scheme. Experiment results show that the proposed model sharing scheme
outperforms the individual model in terms of computing resource consumption and
recognition accuracy.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 15:52:09 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Xiong",
"Kai",
""
],
[
"Wang",
"Zhihong",
""
],
[
"Leng",
"Supeng",
""
],
[
"He",
"Jianhua",
""
]
] | TITLE: A Digital Twin Empowered Lightweight Model Sharing Scheme for
Multi-Robot Systems
ABSTRACT: Multi-robot system for manufacturing is an Industry Internet of Things (IIoT)
paradigm with significant operational cost savings and productivity
improvement, where Unmanned Aerial Vehicles (UAVs) are employed to control and
implement collaborative productions without human intervention. This
mission-critical system relies on 3-Dimension (3-D) scene recognition to
improve operation accuracy in the production line and autonomous piloting.
However, implementing 3-D point cloud learning, such as Pointnet, is
challenging due to limited sensing and computing resources equipped with UAVs.
Therefore, we propose a Digital Twin (DT) empowered Knowledge Distillation (KD)
method to generate several lightweight learning models and select the optimal
model to deploy on UAVs. With a digital replica of the UAVs preserved at the
edge server, the DT system controls the model sharing network topology and
learning model structure to improve recognition accuracy further. Moreover, we
employ network calculus to formulate and solve the model sharing configuration
problem toward minimal resource consumption, as well as convergence. Simulation
experiments are conducted over a popular point cloud dataset to evaluate the
proposed scheme. Experiment results show that the proposed model sharing scheme
outperforms the individual model in terms of computing resource consumption and
recognition accuracy.
| no_new_dataset | 0.707196 |
2305.02219 | Timothy Castiglia Mr. | Timothy Castiglia, Yi Zhou, Shiqiang Wang, Swanand Kadhe, Nathalie
Baracaldo, Stacy Patterson | LESS-VFL: Communication-Efficient Feature Selection for Vertical
Federated Learning | Published in ICML 2023 | null | null | null | cs.LG cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose LESS-VFL, a communication-efficient feature selection method for
distributed systems with vertically partitioned data. We consider a system of a
server and several parties with local datasets that share a sample ID space but
have different feature sets. The parties wish to collaboratively train a model
for a prediction task. As part of the training, the parties wish to remove
unimportant features in the system to improve generalization, efficiency, and
explainability. In LESS-VFL, after a short pre-training period, the server
optimizes its part of the global model to determine the relevant outputs from
party models. This information is shared with the parties to then allow local
feature selection without communication. We analytically prove that LESS-VFL
removes spurious features from model training. We provide extensive empirical
evidence that LESS-VFL can achieve high accuracy and remove spurious features
at a fraction of the communication cost of other feature selection approaches.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 15:56:47 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Castiglia",
"Timothy",
""
],
[
"Zhou",
"Yi",
""
],
[
"Wang",
"Shiqiang",
""
],
[
"Kadhe",
"Swanand",
""
],
[
"Baracaldo",
"Nathalie",
""
],
[
"Patterson",
"Stacy",
""
]
] | TITLE: LESS-VFL: Communication-Efficient Feature Selection for Vertical
Federated Learning
ABSTRACT: We propose LESS-VFL, a communication-efficient feature selection method for
distributed systems with vertically partitioned data. We consider a system of a
server and several parties with local datasets that share a sample ID space but
have different feature sets. The parties wish to collaboratively train a model
for a prediction task. As part of the training, the parties wish to remove
unimportant features in the system to improve generalization, efficiency, and
explainability. In LESS-VFL, after a short pre-training period, the server
optimizes its part of the global model to determine the relevant outputs from
party models. This information is shared with the parties to then allow local
feature selection without communication. We analytically prove that LESS-VFL
removes spurious features from model training. We provide extensive empirical
evidence that LESS-VFL can achieve high accuracy and remove spurious features
at a fraction of the communication cost of other feature selection approaches.
| no_new_dataset | 0.709831 |
2305.02260 | Ante Kapetanovic | Ante Kapetanovic, Dragan Poljak, Kun Li | Standardized Benchmark Dataset for Localized Exposure to a Realistic
Source at 10$-$90 GHz | 6 pages, 3 figures, in proceedings of BioEM2023 | null | null | null | physics.med-ph cs.LG eess.IV | http://creativecommons.org/licenses/by-sa/4.0/ | The lack of freely available standardized datasets represents an aggravating
factor during the development and testing the performance of novel
computational techniques in exposure assessment and dosimetry research. This
hinders progress as researchers are required to generate numerical data (field,
power and temperature distribution) anew using simulation software for each
exposure scenario. Other than being time consuming, this approach is highly
susceptible to errors that occur during the configuration of the
electromagnetic model. To address this issue, in this paper, the limited
available data on the incident power density and resultant maximum temperature
rise on the skin surface considering various steady-state exposure scenarios at
10$-$90 GHz have been statistically modeled. The synthetic data have been
sampled from the fitted statistical multivariate distribution with respect to
predetermined dosimetric constraints. We thus present a comprehensive and
open-source dataset compiled of the high-fidelity numerical data considering
various exposures to a realistic source. Furthermore, different surrogate
models for predicting maximum temperature rise on the skin surface were fitted
based on the synthetic dataset. All surrogate models were tested on the
originally available data where satisfactory predictive performance has been
demonstrated. A simple technique of combining quadratic polynomial and
tensor-product spline surrogates, each operating on its own cluster of data,
has achieved the lowest mean absolute error of 0.058 {\deg}C. Therefore,
overall experimental results indicate the validity of the proposed synthetic
dataset.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 16:48:01 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Kapetanovic",
"Ante",
""
],
[
"Poljak",
"Dragan",
""
],
[
"Li",
"Kun",
""
]
] | TITLE: Standardized Benchmark Dataset for Localized Exposure to a Realistic
Source at 10$-$90 GHz
ABSTRACT: The lack of freely available standardized datasets represents an aggravating
factor during the development and testing the performance of novel
computational techniques in exposure assessment and dosimetry research. This
hinders progress as researchers are required to generate numerical data (field,
power and temperature distribution) anew using simulation software for each
exposure scenario. Other than being time consuming, this approach is highly
susceptible to errors that occur during the configuration of the
electromagnetic model. To address this issue, in this paper, the limited
available data on the incident power density and resultant maximum temperature
rise on the skin surface considering various steady-state exposure scenarios at
10$-$90 GHz have been statistically modeled. The synthetic data have been
sampled from the fitted statistical multivariate distribution with respect to
predetermined dosimetric constraints. We thus present a comprehensive and
open-source dataset compiled of the high-fidelity numerical data considering
various exposures to a realistic source. Furthermore, different surrogate
models for predicting maximum temperature rise on the skin surface were fitted
based on the synthetic dataset. All surrogate models were tested on the
originally available data where satisfactory predictive performance has been
demonstrated. A simple technique of combining quadratic polynomial and
tensor-product spline surrogates, each operating on its own cluster of data,
has achieved the lowest mean absolute error of 0.058 {\deg}C. Therefore,
overall experimental results indicate the validity of the proposed synthetic
dataset.
| new_dataset | 0.636777 |
2305.02292 | Soheila Hatami | Soheila Hatami, Majid Sadedel, Farideh Jamali | Iranian License Plate Recognition Using a Reliable Deep Learning
Approach | Under Review in Scientia Iranica Journal | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The issue of Automatic License Plate Recognition (ALPR) has been one of the
most challenging issues in recent years. Weather conditions, camera angle of
view, lighting conditions, different characters written on license plates, and
many other factors are among the challenges for the issue of ALPR. Given the
advances that have been made in recent years in the field of deep neural
networks, some types of neural networks and models based on them can be used to
perform the task of Iranian license plate recognition. In the proposed method
presented in this paper, the license plate recognition is done in two steps.
The first step is to detect the rectangles of the license plates from the input
image. In the second step, these license plates are cropped from the image and
their characters are recognized. For the first step, 3065 images including
license plates and for the second step, 3364 images including characters of
license plates have been prepared and considered as the desired datasets. In
the first step, license plates are detected using the YOLOv4-tiny model, which
is based on Convolutional Neural Network (CNN). In the next step, the
characters of these license plates are recognized using Convolutional Recurrent
Neural Network (CRNN), and Connectionist Temporal Classification (CTC). In the
second step, there is no need to segment and label the characters separately,
only one string of numbers and letters is enough for the labels.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 17:34:10 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Hatami",
"Soheila",
""
],
[
"Sadedel",
"Majid",
""
],
[
"Jamali",
"Farideh",
""
]
] | TITLE: Iranian License Plate Recognition Using a Reliable Deep Learning
Approach
ABSTRACT: The issue of Automatic License Plate Recognition (ALPR) has been one of the
most challenging issues in recent years. Weather conditions, camera angle of
view, lighting conditions, different characters written on license plates, and
many other factors are among the challenges for the issue of ALPR. Given the
advances that have been made in recent years in the field of deep neural
networks, some types of neural networks and models based on them can be used to
perform the task of Iranian license plate recognition. In the proposed method
presented in this paper, the license plate recognition is done in two steps.
The first step is to detect the rectangles of the license plates from the input
image. In the second step, these license plates are cropped from the image and
their characters are recognized. For the first step, 3065 images including
license plates and for the second step, 3364 images including characters of
license plates have been prepared and considered as the desired datasets. In
the first step, license plates are detected using the YOLOv4-tiny model, which
is based on Convolutional Neural Network (CNN). In the next step, the
characters of these license plates are recognized using Convolutional Recurrent
Neural Network (CRNN), and Connectionist Temporal Classification (CTC). In the
second step, there is no need to segment and label the characters separately,
only one string of numbers and letters is enough for the labels.
| no_new_dataset | 0.712426 |
2305.02296 | Nikita Karaev | Nikita Karaev, Ignacio Rocco, Benjamin Graham, Natalia Neverova,
Andrea Vedaldi, Christian Rupprecht | DynamicStereo: Consistent Dynamic Depth from Stereo Videos | CVPR 2023; project page available at
https://dynamic-stereo.github.io/ | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the problem of reconstructing a dynamic scene observed from a
stereo camera. Most existing methods for depth from stereo treat different
stereo frames independently, leading to temporally inconsistent depth
predictions. Temporal consistency is especially important for immersive AR or
VR scenarios, where flickering greatly diminishes the user experience. We
propose DynamicStereo, a novel transformer-based architecture to estimate
disparity for stereo videos. The network learns to pool information from
neighboring frames to improve the temporal consistency of its predictions. Our
architecture is designed to process stereo videos efficiently through divided
attention layers. We also introduce Dynamic Replica, a new benchmark dataset
containing synthetic videos of people and animals in scanned environments,
which provides complementary training and evaluation data for dynamic stereo
closer to real applications than existing datasets. Training with this dataset
further improves the quality of predictions of our proposed DynamicStereo as
well as prior methods. Finally, it acts as a benchmark for consistent stereo
methods.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 17:40:49 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Karaev",
"Nikita",
""
],
[
"Rocco",
"Ignacio",
""
],
[
"Graham",
"Benjamin",
""
],
[
"Neverova",
"Natalia",
""
],
[
"Vedaldi",
"Andrea",
""
],
[
"Rupprecht",
"Christian",
""
]
] | TITLE: DynamicStereo: Consistent Dynamic Depth from Stereo Videos
ABSTRACT: We consider the problem of reconstructing a dynamic scene observed from a
stereo camera. Most existing methods for depth from stereo treat different
stereo frames independently, leading to temporally inconsistent depth
predictions. Temporal consistency is especially important for immersive AR or
VR scenarios, where flickering greatly diminishes the user experience. We
propose DynamicStereo, a novel transformer-based architecture to estimate
disparity for stereo videos. The network learns to pool information from
neighboring frames to improve the temporal consistency of its predictions. Our
architecture is designed to process stereo videos efficiently through divided
attention layers. We also introduce Dynamic Replica, a new benchmark dataset
containing synthetic videos of people and animals in scanned environments,
which provides complementary training and evaluation data for dynamic stereo
closer to real applications than existing datasets. Training with this dataset
further improves the quality of predictions of our proposed DynamicStereo as
well as prior methods. Finally, it acts as a benchmark for consistent stereo
methods.
| new_dataset | 0.715194 |
2305.02307 | Mengyun Shi | Mengyun Shi, Serge Belongie, Claire Cardie | Fashionpedia-Taste: A Dataset towards Explaining Human Fashion Taste | null | null | null | null | cs.CV cs.AI cs.DB | http://creativecommons.org/licenses/by/4.0/ | Existing fashion datasets do not consider the multi-facts that cause a
consumer to like or dislike a fashion image. Even two consumers like a same
fashion image, they could like this image for total different reasons. In this
paper, we study the reason why a consumer like a certain fashion image. Towards
this goal, we introduce an interpretability dataset, Fashionpedia-taste,
consist of rich annotation to explain why a subject like or dislike a fashion
image from the following 3 perspectives: 1) localized attributes; 2) human
attention; 3) caption. Furthermore, subjects are asked to provide their
personal attributes and preference on fashion, such as personality and
preferred fashion brands. Our dataset makes it possible for researchers to
build computational models to fully understand and interpret human fashion
taste from different humanistic perspectives and modalities.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 17:54:50 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Shi",
"Mengyun",
""
],
[
"Belongie",
"Serge",
""
],
[
"Cardie",
"Claire",
""
]
] | TITLE: Fashionpedia-Taste: A Dataset towards Explaining Human Fashion Taste
ABSTRACT: Existing fashion datasets do not consider the multi-facts that cause a
consumer to like or dislike a fashion image. Even two consumers like a same
fashion image, they could like this image for total different reasons. In this
paper, we study the reason why a consumer like a certain fashion image. Towards
this goal, we introduce an interpretability dataset, Fashionpedia-taste,
consist of rich annotation to explain why a subject like or dislike a fashion
image from the following 3 perspectives: 1) localized attributes; 2) human
attention; 3) caption. Furthermore, subjects are asked to provide their
personal attributes and preference on fashion, such as personality and
preferred fashion brands. Our dataset makes it possible for researchers to
build computational models to fully understand and interpret human fashion
taste from different humanistic perspectives and modalities.
| new_dataset | 0.714416 |
2305.02310 | Alex Trevithick | Alex Trevithick, Matthew Chan, Michael Stengel, Eric R. Chan, Chao
Liu, Zhiding Yu, Sameh Khamis, Manmohan Chandraker, Ravi Ramamoorthi, Koki
Nagano | Real-Time Radiance Fields for Single-Image Portrait View Synthesis | Project page: https://research.nvidia.com/labs/nxp/lp3d/ | null | null | null | cs.CV cs.AI cs.GR cs.LG | http://creativecommons.org/licenses/by/4.0/ | We present a one-shot method to infer and render a photorealistic 3D
representation from a single unposed image (e.g., face portrait) in real-time.
Given a single RGB input, our image encoder directly predicts a canonical
triplane representation of a neural radiance field for 3D-aware novel view
synthesis via volume rendering. Our method is fast (24 fps) on consumer
hardware, and produces higher quality results than strong GAN-inversion
baselines that require test-time optimization. To train our triplane encoder
pipeline, we use only synthetic data, showing how to distill the knowledge from
a pretrained 3D GAN into a feedforward encoder. Technical contributions include
a Vision Transformer-based triplane encoder, a camera data augmentation
strategy, and a well-designed loss function for synthetic data training. We
benchmark against the state-of-the-art methods, demonstrating significant
improvements in robustness and image quality in challenging real-world
settings. We showcase our results on portraits of faces (FFHQ) and cats (AFHQ),
but our algorithm can also be applied in the future to other categories with a
3D-aware image generator.
| [
{
"version": "v1",
"created": "Wed, 3 May 2023 17:56:01 GMT"
}
] | 2023-05-04T00:00:00 | [
[
"Trevithick",
"Alex",
""
],
[
"Chan",
"Matthew",
""
],
[
"Stengel",
"Michael",
""
],
[
"Chan",
"Eric R.",
""
],
[
"Liu",
"Chao",
""
],
[
"Yu",
"Zhiding",
""
],
[
"Khamis",
"Sameh",
""
],
[
"Chandraker",
"Manmohan",
""
],
[
"Ramamoorthi",
"Ravi",
""
],
[
"Nagano",
"Koki",
""
]
] | TITLE: Real-Time Radiance Fields for Single-Image Portrait View Synthesis
ABSTRACT: We present a one-shot method to infer and render a photorealistic 3D
representation from a single unposed image (e.g., face portrait) in real-time.
Given a single RGB input, our image encoder directly predicts a canonical
triplane representation of a neural radiance field for 3D-aware novel view
synthesis via volume rendering. Our method is fast (24 fps) on consumer
hardware, and produces higher quality results than strong GAN-inversion
baselines that require test-time optimization. To train our triplane encoder
pipeline, we use only synthetic data, showing how to distill the knowledge from
a pretrained 3D GAN into a feedforward encoder. Technical contributions include
a Vision Transformer-based triplane encoder, a camera data augmentation
strategy, and a well-designed loss function for synthetic data training. We
benchmark against the state-of-the-art methods, demonstrating significant
improvements in robustness and image quality in challenging real-world
settings. We showcase our results on portraits of faces (FFHQ) and cats (AFHQ),
but our algorithm can also be applied in the future to other categories with a
3D-aware image generator.
| no_new_dataset | 0.710785 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.