Deep-fake-Forensics-Challenge/utils.py at main · Goodnight77/Deep-fake-Forensics-Challenge · GitHub

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

import numpy as np
import tensorflow as tf
from tensorflow.keras.preprocessing.image import load_img, img_to_array
import tensorflow as tf
from tqdm import tqdm
from tensorflow.keras import backend as K
from tensorflow import keras
from tensorflow.keras import layers

#this is the set of learnable filters
class BayarConv2d(tf.keras.layers.Layer):
    def __init__(self, in_channels, out_channels, kernel_size=5, stride=1, padding=0):
        super(BayarConv2d, self).__init__()
        self.in_channels = in_channels
        self.out_channels = out_channels
        self.kernel_size = kernel_size
        self.stride = stride
        self.padding = padding
        self.minus1 = tf.ones((self.in_channels, self.out_channels, 1)) * -1.000

        # only (kernel_size ** 2 - 1) trainable params as the center element is always -1
        self.kernel = self.add_weight(shape=(self.in_channels, self.out_channels, kernel_size ** 2 - 1),
                                      initializer='random_normal',
                                      trainable=True)

    def bayarConstraint(self):
        kernel_permuted = tf.transpose(self.kernel, perm=[2, 0, 1])
        kernel_sum = tf.reduce_sum(kernel_permuted, axis=0)
        ctr = self.kernel_size ** 2 // 2
        real_kernel = tf.concat([self.kernel[:, :, :ctr], self.minus1, self.kernel[:, :, ctr:]], axis=2)
        real_kernel = tf.reshape(real_kernel, (self.out_channels, self.in_channels, self.kernel_size, self.kernel_size))
        return real_kernel

    def call(self, x):
        x = tf.nn.conv2d(x, self.bayarConstraint(), strides=self.stride, padding='SAME')
        return x

def add_random_boxes(img,n_k,size=32):
    h,w = size,size
    img = np.asarray(img)
    img_size = img.shape[1]
    boxes = []
    for k in range(n_k):
        y,x = np.random.randint(0,img_size-w,(2,))
        img[y:y+h,x:x+w] = 0
        boxes.append((x,y,h,w))
    return img
class DataGeneratorA(tf.keras.utils.Sequence):
    def __init__(self, data, batch_size=32, shuffle=True, augment=True):
        self.data = data
        self.batch_size = batch_size
        self.shuffle = shuffle
        self.augment = augment
        self.on_epoch_end()
        self.Image_gen = tf.keras.preprocessing.image.ImageDataGenerator(
            brightness_range=[0.2, 1.8]
        )
        self.iimage_gen = tf.keras.preprocessing.image.ImageDataGenerator(
            brightness_range=[3, 4]
        )
        if self.augment:
            self.image_datagen = tf.keras.preprocessing.image.ImageDataGenerator(
                rotation_range=45,
                width_shift_range=0.2,
                height_shift_range=0.2,
                shear_range=0.2,
                zoom_range=0.2,
                horizontal_flip=True,
                fill_mode='nearest',
            )

    def __len__(self):
        return int(np.ceil(len(self.data) / self.batch_size))

    def __getitem__(self, index):
        batch_data = self.data[index * self.batch_size: (index + 1) * self.batch_size]

        images = []
        masks = []
        labels = []

        for _, row in batch_data.iterrows():
            image = load_img(row['image'], target_size=(256, 256))
            image = img_to_array(image)
            if self.augment==True:
                image = self.Image_gen.random_transform(image)
            edge = load_img(row['edge'], target_size=(256, 256))
            edge = img_to_array(edge)
            edge/=255.0

            mask = load_img(row['mask'], target_size=(256, 256))
            image/=255.0
            mask = img_to_array(mask) / 255.0
            modified=image*edge
            aug=self.iimage_gen.random_transform(modified*255.0)
            final=image-modified+aug/255.0
            augmented = add_random_boxes(np.concatenate([final, mask], axis=-1),3,70)
            final, mask = np.split(augmented, 2, axis=-1)


            if self.augment:
                augmented = self.image_datagen.random_transform(np.concatenate([final, mask], axis=-1))
                final, mask = np.split(augmented, 2, axis=-1)
                mask=tf.image.rgb_to_grayscale(mask)

            images.append(final)
            masks.append(mask)
            labels.append(row['class'])

        images = np.array(images)
        masks = np.array(masks)
        labels = np.array(labels)

        return images, [masks, labels]

    def on_epoch_end(self):
        if self.shuffle:
            self.data = self.data.sample(frac=1)


class DataGeneratorNA(tf.keras.utils.Sequence):
    def __init__(self, data, batch_size=32, shuffle=True):
        self.data = data
        self.batch_size = batch_size
        self.shuffle = shuffle
        self.on_epoch_end()

    def __len__(self):
        return int(np.ceil(len(self.data) / self.batch_size))

    def __getitem__(self, index):
        batch_data = self.data[index * self.batch_size : (index + 1) * self.batch_size]

        images = []
        masks = []
        labels = []

        for _, row in batch_data.iterrows():
            image = load_img(row['image'], target_size=(256, 256))
            image = img_to_array(image) / 255.0
            images.append(image)

            mask = load_img(row['mask'], target_size=(256, 256), color_mode='grayscale')
            mask = img_to_array(mask) / 255.0
            masks.append(mask)

            labels.append(row['class'])

        images = np.array(images)
        masks = np.array(masks)
        labels = np.array(labels)

        return images, [masks, labels]

    def on_epoch_end(self):
        if self.shuffle:
            self.data = self.data.sample(frac=1)