File size: 6,728 Bytes
d4004bc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6e1c8e6
 
d4004bc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6e1c8e6
d4004bc
6e1c8e6
d4004bc
 
 
 
 
6e1c8e6
d4004bc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6e1c8e6
d4004bc
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
"""
Copyright $today.year LY Corporation

LY Corporation licenses this file to you 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:

  https://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.
"""
import os
import torch
import subprocess
import gradio as gr
import librosa
from tqdm import tqdm
from lighthouse.models import *

# use GPU if available
device = "cuda" if torch.cuda.is_available() else "cpu"
MODEL_NAMES = ['qd_detr']
FEATURES = ['clap']
TOPK_MOMENT = 5
sample_path = "sample_data/1a-ODBWMUAE.wav"
sample_query = "Water cascades down from a waterfall."

"""
Helper functions
"""
def load_pretrained_weights():
    file_urls = []
    for model_name in MODEL_NAMES:
        for feature in FEATURES:
            file_urls.append(
                "https://zenodo.org/records/13961029/files/{}_{}_clotho-moment.ckpt".format(feature, model_name)
            )
    for file_url in tqdm(file_urls):
        if not os.path.exists('gradio_demo/weights/' + os.path.basename(file_url)):
            command = 'wget -P gradio_demo/weights/ {}'.format(file_url)
            subprocess.run(command, shell=True)

    return file_urls

def flatten(array2d):
    list1d = []
    for elem in array2d:
        list1d += elem
    return list1d

"""
Model initialization
"""
load_pretrained_weights()
model = QDDETRPredictor('gradio_demo/weights/clap_qd_detr_clotho-moment.ckpt', device=device, feature_name='clap')
loaded_audio = None

"""
Gradio functions
"""
def audio_upload(audio):
    global loaded_audio
    if audio is None:
        loaded_audio = None
        yield gr.update(value="Removed the audio", visible=True)
    else:
        yield gr.update(value="Processing the audio. Wait for a minute...", visible=True)
        audio_feats = model.encode_audio(audio)
        loaded_audio = audio_feats
        yield gr.update(value="Finished audio processing!", visible=True)

def model_load(radio):
    if radio is not None:
        yield gr.update(value="Loading new model. Wait for a minute...", visible=True)
        global model
        feature, model_name = radio.split('+')
        feature, model_name = feature.strip(), model_name.strip()

        if model_name == 'qd_detr':
            model_class = QDDETRPredictor
        else:
            raise gr.Error("Select from the models")

        model = model_class('gradio_demo/weights/{}_{}_clotho-moment.ckpt'.format(feature, model_name),
                            device=device, feature_name='{}'.format(feature))
        yield gr.update(value="Model loaded: {}".format(radio), visible=True)

def predict(textbox, line, gallery):
    global loaded_audio
    if loaded_audio is None:
        raise gr.Error('Upload the audio before pushing the `Retrieve moment` button.')
    else:
        prediction = model.predict(textbox, loaded_audio)
        mr_results = prediction['pred_relevant_windows']

        buttons = []
        for i, pred in enumerate(mr_results[:TOPK_MOMENT]):
            buttons.append(gr.Button(value='moment {}: [{}, {}] Score: {}'.format(i+1, pred[0], pred[1], pred[2]), visible=True))

        return buttons


def show_trimmed_audio(audio, button):
    s, sr = librosa.load(audio, sr=None)
    _seconds = button.split(': [')[1].split(']')[0].split(', ')
    start_sec = float(_seconds[0])
    end_sec = float(_seconds[1])
    start_frame = int(start_sec * sr)
    end_frame = int(end_sec * sr)

    return gr.Audio((sr, s[start_frame:end_frame]), interactive=False, visible=True)


def main():
    title = """# Audio Moment Retrieval Demo"""

    with gr.Blocks(theme=gr.themes.Soft()) as demo:
        gr.Markdown(title)

        with gr.Row():
            with gr.Column():
                with gr.Group():
                    gr.Markdown("## Model selection")
                    radio_list = flatten([["{} + {}".format(feature, model_name) for model_name in MODEL_NAMES] for feature in FEATURES])
                    radio = gr.Radio(radio_list, label="models", value="clap + qd_detr", info="Which model do you want to use?")
                    load_status_text = gr.Textbox(label='Model load status', value='Model loaded: clap + qd_detr')

                with gr.Group():
                    gr.Markdown("## Audio and query")
                    audio_input = gr.Audio(sample_path, type='filepath')
                    output = gr.Textbox(label='Audio processing progress')
                    query_input = gr.Textbox(sample_query, label='query')
                    button = gr.Button("Retrieve moment", variant="primary")

            with gr.Column():
                with gr.Group():
                    gr.Markdown("## Retrieved moments")
                    gr.Markdown("Click on the moment button to listen to the trimmed audio.")
                    button_1 = gr.Button(value='moment 1', visible=False, elem_id='result_0')
                    button_2 = gr.Button(value='moment 2', visible=False, elem_id='result_1')
                    button_3 = gr.Button(value='moment 3', visible=False, elem_id='result_2')
                    button_4 = gr.Button(value='moment 4', visible=False, elem_id='result_3')
                    button_5 = gr.Button(value='moment 5', visible=False, elem_id='result_4')
                    result = gr.Audio(None, label='Trimmed audio', interactive=False, visible=False)

                    button_1.click(show_trimmed_audio, inputs=[audio_input, button_1], outputs=[result])
                    button_2.click(show_trimmed_audio, inputs=[audio_input, button_2], outputs=[result])
                    button_3.click(show_trimmed_audio, inputs=[audio_input, button_3], outputs=[result])
                    button_4.click(show_trimmed_audio, inputs=[audio_input, button_4], outputs=[result])
                    button_5.click(show_trimmed_audio, inputs=[audio_input, button_5], outputs=[result])

                audio_input.change(audio_upload, inputs=[audio_input], outputs=output)
                radio.select(model_load, inputs=[radio], outputs=load_status_text)
                
                button.click(predict, 
                            inputs=[query_input], 
                            outputs=[button_1, button_2, button_3, button_4, button_5])
        demo.load(audio_upload, inputs=[audio_input], outputs=output)

    demo.launch()

if __name__ == "__main__":
    main()