File size: 4,388 Bytes
fb59cb8
 
 
 
 
 
 
 
e9c5f95
fb59cb8
 
 
 
 
 
 
 
f5184a5
 
 
 
fb59cb8
f5184a5
e9c5f95
 
 
fb59cb8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e9c5f95
 
 
 
fb59cb8
e9c5f95
fb59cb8
 
e9c5f95
 
 
fb59cb8
 
e9c5f95
 
 
 
 
 
fb59cb8
 
e9c5f95
 
 
 
 
 
 
 
 
 
 
fb59cb8
 
 
 
 
 
 
 
 
 
 
 
18988d2
fb59cb8
 
 
 
 
 
 
 
 
 
 
e9c5f95
fb59cb8
 
 
e9c5f95
 
fb59cb8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f5184a5
 
 
 
fb59cb8
 
 
 
 
 
 
 
 
 
 
 
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
#!/usr/bin/env python

from __future__ import annotations

import argparse
import functools
import os
import pathlib
import tarfile

import deepdanbooru as dd
import gradio as gr
import huggingface_hub
import numpy as np
import PIL.Image
import tensorflow as tf

ORIGINAL_REPO_URL = 'https://github.com/KichangKim/DeepDanbooru'
TITLE = 'KichangKim/DeepDanbooru'
DESCRIPTION = f'A demo for {ORIGINAL_REPO_URL}'
ARTICLE = None

TOKEN = os.environ['TOKEN']
MODEL_REPO = 'hysts/DeepDanbooru'
MODEL_FILENAME = 'model-resnet_custom_v3.h5'
LABEL_FILENAME = 'tags.txt'


def parse_args() -> argparse.Namespace:
    parser = argparse.ArgumentParser()
    parser.add_argument('--score-slider-step', type=float, default=0.05)
    parser.add_argument('--score-threshold', type=float, default=0.5)
    parser.add_argument('--theme', type=str, default='dark-grass')
    parser.add_argument('--live', action='store_true')
    parser.add_argument('--share', action='store_true')
    parser.add_argument('--port', type=int)
    parser.add_argument('--disable-queue',
                        dest='enable_queue',
                        action='store_false')
    parser.add_argument('--allow-flagging', type=str, default='never')
    parser.add_argument('--allow-screenshot', action='store_true')
    return parser.parse_args()


def load_sample_image_paths() -> list[pathlib.Path]:
    image_dir = pathlib.Path('images')
    if not image_dir.exists():
        dataset_repo = 'hysts/sample-images-TADNE'
        path = huggingface_hub.hf_hub_download(dataset_repo,
                                               'images.tar.gz',
                                               repo_type='dataset',
                                               use_auth_token=TOKEN)
        with tarfile.open(path) as f:
            f.extractall()
    return sorted(image_dir.glob('*'))


def load_model() -> tf.keras.Model:
    path = huggingface_hub.hf_hub_download(MODEL_REPO,
                                           MODEL_FILENAME,
                                           use_auth_token=TOKEN)
    model = tf.keras.models.load_model(path)
    return model


def load_labels() -> list[str]:
    path = huggingface_hub.hf_hub_download(MODEL_REPO,
                                           LABEL_FILENAME,
                                           use_auth_token=TOKEN)
    with open(path) as f:
        labels = [line.strip() for line in f.readlines()]
    return labels


def predict(image: PIL.Image.Image, score_threshold: float,
            model: tf.keras.Model, labels: list[str]) -> dict[str, float]:
    _, height, width, _ = model.input_shape
    image = np.asarray(image)
    image = tf.image.resize(image,
                            size=(height, width),
                            method=tf.image.ResizeMethod.AREA,
                            preserve_aspect_ratio=True)
    image = image.numpy()
    image = dd.image.transform_and_pad_image(image, width, height)
    image = image / 255.
    probs = model.predict(image[None, ...])[0]
    probs = probs.astype(float)
    res = dict()
    for prob, label in zip(probs.tolist(), labels):
        if prob < score_threshold:
            continue
        res[label] = prob
    return res


def main():
    gr.close_all()

    args = parse_args()

    image_paths = load_sample_image_paths()
    examples = [[path.as_posix(), args.score_threshold]
                for path in image_paths]

    model = load_model()
    labels = load_labels()

    func = functools.partial(predict, model=model, labels=labels)
    func = functools.update_wrapper(func, predict)

    gr.Interface(
        func,
        [
            gr.inputs.Image(type='pil', label='Input'),
            gr.inputs.Slider(0,
                             1,
                             step=args.score_slider_step,
                             default=args.score_threshold,
                             label='Score Threshold'),
        ],
        gr.outputs.Label(label='Output'),
        examples=examples,
        title=TITLE,
        description=DESCRIPTION,
        article=ARTICLE,
        theme=args.theme,
        allow_screenshot=args.allow_screenshot,
        allow_flagging=args.allow_flagging,
        live=args.live,
    ).launch(
        enable_queue=args.enable_queue,
        server_port=args.port,
        share=args.share,
    )


if __name__ == '__main__':
    main()