HyperForensics-plus-plus.py

# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""HyperForensics++ dataset"""


import csv
import json
import os
import numpy as np

import datasets

_CITATION = """\
@InProceedings{hyperforensics:dataset,
title={HyperForensics++: Toward Adversarial Perturbed and Object Replacement in Hyperspectral Imaging Dataset},
author={Chih-Chung Hsu, Chia-Ming Lee, Min-Tzo Ko, En-Chao Liu, Yi-Ching Cheng, Ming-Ching Chang},
year={2025}
}
"""

_DESCRIPTION = """\
The HyperForensics++ dataset is an advanced benchmark designed for hyperspectral image (HSI) forgery detection.
It builds upon the foundational HyperForensics dataset by introducing new manipulation scenarios and enhanced techniques.
"""

_HOMEPAGE = "https://huggingface.co/datasets/OtoroLin/HyperForensics-plus-plus"

# TODO: Add the licence for the dataset here if you can find it
_LICENSE = ""

_URL = "https://huggingface.co/datasets/OtoroLin/HyperForensics-plus-plus/resolve/main/data.tar.gz"


class HyperForensicsPlusPlus(datasets.GeneratorBasedBuilder):
    """HyperForensics++ dataset"""

    VERSION = datasets.Version("1.1.0")

    # If you need to make complex sub-parts in the datasets with configurable options
    # You can create your own builder configuration class to store attribute, inheriting from datasets.BuilderConfig
    # BUILDER_CONFIG_CLASS = MyBuilderConfig

    # You will be able to load one or the other configurations in the following list with
    # data = datasets.load_dataset('my_dataset', 'first_domain')
    # data = datasets.load_dataset('my_dataset', 'second_domain')
    BUILDER_CONFIGS = [
        datasets.BuilderConfig(
            name="data",
            version=VERSION,
            description="Full dataset with all the HSI in npy format",),
        # datasets.BuilderConfig(name="second_domain", version=VERSION, description="This part of my dataset covers a second domain"),
    ]

    DEFAULT_CONFIG_NAME = "data"  # It's not mandatory to have a default configuration. Just use one if it make sense.
    def __init__(self, **kwargs):
        # You can add any custom arguments here that you want to pass to the builder
        # They will be passed to the constructor of the parent class
        super(HyperForensicsPlusPlus, self).__init__(writer_batch_size=4, **kwargs)

    def _info(self):
        features = datasets.Features(
            {
                "origin": datasets.Array3D(dtype="int16", shape=(256, 256, 172)), # The original HSI
                "label": datasets.Value("string"), # The label of the image
                "forgery": datasets.Array3D(dtype="int16", shape=(256, 256, 172)), # The HSI after forgery
                "method": datasets.Value("string"), # The forgery method used
                # The bounding box of the forgery area, in the format [x1, x2, y1, y2, z1, y2]
                "bbox": datasets.Sequence(feature=datasets.Value(dtype='int16'), length=6)
            }
        )
 
        return datasets.DatasetInfo(
            # This is the description that will appear on the datasets page.
            description=_DESCRIPTION,
            # This defines the different columns of the dataset and their types
            features=features,  # Here we define them above because they are different between the two configurations
            # If there's a common (input, target) tuple from the features, uncomment supervised_keys line below and
            # specify them. They'll be used if as_supervised=True in builder.as_dataset.
            # supervised_keys=("sentence", "label"),
            # Homepage of the dataset for documentation
            homepage=_HOMEPAGE,
            # License for the dataset if available
            license=_LICENSE,
            # Citation for the dataset
            citation=_CITATION,
        )

    def _split_generators(self, dl_manager):
        # dl_manager is a datasets.download.DownloadManager that can be used to download and extract URLS
        # It can accept any type or nested list/dict and will give back the same structure with the url replaced with path to local files.
        # By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
        urls = _URL
        data_dir = dl_manager.download_and_extract(urls)


        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                # These kwargs will be passed to _generate_examples
                gen_kwargs={
                    "filepath": data_dir,
                    "split": "train",
                },
            ),
            datasets.SplitGenerator(
                name=datasets.Split.VALIDATION,
                # These kwargs will be passed to _generate_examples
                gen_kwargs={
                    "filepath": data_dir,
                    "split": "validation",
                },
            ),
            #datasets.SplitGenerator(
            #    name=datasets.Split.TEST,
            #    # These kwargs will be passed to _generate_examples
            #    gen_kwargs={
            #        "filepath": data_dir,
            #        "split": "testing"
            #    },
            #),
        ]

    # method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
    def _generate_examples(self, filepath, split):
        filepath = os.path.join(filepath, "data_testing")
        with open(os.path.join(filepath, "metadata.jsonl"), encoding="utf-8") as f:
            metadata = json.load(f)  # Load the nested JSON object (train, validation, testing)

            # Select the appropriate split (train, validation, or testing)
            records = metadata[split]

            for key, record in enumerate(records):
                file_prefix = record["file_prefix"]
                label = record["label"]
                bbox = record["bbox"]

                # Construct paths for the origin and forgery files
                origin_path = os.path.join(
                    filepath, "ADMM-ADAM", "config0", f"{file_prefix}_inpaint_result(0).npy"
                )
                forgery_path = os.path.join(
                    filepath, "ADMM-ADAM", "config0", f"{file_prefix}_inpaint_result(0).npy"
                )

                # Load the .npy files as images
                origin_image = np.load(origin_path) #np.load(origin_path)
                forgery_image = np.load(forgery_path) #np.load(forgery_path)
                # Yield the example
                yield key, {
                    "origin": origin_image,
                    "label": label,
                    "forgery": forgery_image,
                    "method": "ADMM-ADAM",  # Hardcoded for now; can be dynamic if needed
                    "bbox": bbox,
                }