EfficientAD/common.py at main · apple024/EfficientAD · 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
#!/usr/bin/python
# -*- coding: utf-8 -*-
from torch import nn
from torchvision.datasets import ImageFolder, DatasetFolder, VisionDataset
from torchvision.io import read_image
from sklearn.metrics import precision_recall_curve
import numpy as np
import os
from PIL import Image

def get_autoencoder(out_channels=384):
    return nn.Sequential(
        # encoder
        nn.Conv2d(in_channels=3, out_channels=32, kernel_size=4, stride=2,
                  padding=1),
        nn.ReLU(inplace=True),
        nn.Conv2d(in_channels=32, out_channels=32, kernel_size=4, stride=2,
                  padding=1),
        nn.ReLU(inplace=True),
        nn.Conv2d(in_channels=32, out_channels=64, kernel_size=4, stride=2,
                  padding=1),
        nn.ReLU(inplace=True),
        nn.Conv2d(in_channels=64, out_channels=64, kernel_size=4, stride=2,
                  padding=1),
        nn.ReLU(inplace=True),
        nn.Conv2d(in_channels=64, out_channels=64, kernel_size=4, stride=2,
                  padding=1),
        nn.ReLU(inplace=True),
        nn.Conv2d(in_channels=64, out_channels=64, kernel_size=8),
        # decoder
        nn.Upsample(size=3, mode='bilinear'),
        nn.Conv2d(in_channels=64, out_channels=64, kernel_size=4, stride=1,
                  padding=2),
        nn.ReLU(inplace=True),
        nn.Dropout(0.2),
        nn.Upsample(size=8, mode='bilinear'),
        nn.Conv2d(in_channels=64, out_channels=64, kernel_size=4, stride=1,
                  padding=2),
        nn.ReLU(inplace=True),
        nn.Dropout(0.2),
        nn.Upsample(size=15, mode='bilinear'),
        nn.Conv2d(in_channels=64, out_channels=64, kernel_size=4, stride=1,
                  padding=2),
        nn.ReLU(inplace=True),
        nn.Dropout(0.2),
        nn.Upsample(size=32, mode='bilinear'),
        nn.Conv2d(in_channels=64, out_channels=64, kernel_size=4, stride=1,
                  padding=2),
        nn.ReLU(inplace=True),
        nn.Dropout(0.2),
        nn.Upsample(size=63, mode='bilinear'),
        nn.Conv2d(in_channels=64, out_channels=64, kernel_size=4, stride=1,
                  padding=2),
        nn.ReLU(inplace=True),
        nn.Dropout(0.2),
        nn.Upsample(size=127, mode='bilinear'),
        nn.Conv2d(in_channels=64, out_channels=64, kernel_size=4, stride=1,
                  padding=2),
        nn.ReLU(inplace=True),
        nn.Dropout(0.2),
        nn.Upsample(size=56, mode='bilinear'),
        nn.Conv2d(in_channels=64, out_channels=64, kernel_size=3, stride=1,
                  padding=1),
        nn.ReLU(inplace=True),
        nn.Conv2d(in_channels=64, out_channels=out_channels, kernel_size=3,
                  stride=1, padding=1)
    )

def get_pdn_small(out_channels=384, padding=False):
    pad_mult = 1 if padding else 0
    return nn.Sequential(
        nn.Conv2d(in_channels=3, out_channels=128, kernel_size=4,
                  padding=3 * pad_mult),
        nn.ReLU(inplace=True),
        nn.AvgPool2d(kernel_size=2, stride=2, padding=1 * pad_mult),
        nn.Conv2d(in_channels=128, out_channels=256, kernel_size=4,
                  padding=3 * pad_mult),
        nn.ReLU(inplace=True),
        nn.AvgPool2d(kernel_size=2, stride=2, padding=1 * pad_mult),
        nn.Conv2d(in_channels=256, out_channels=256, kernel_size=3,
                  padding=1 * pad_mult),
        nn.ReLU(inplace=True),
        nn.Conv2d(in_channels=256, out_channels=out_channels, kernel_size=4)
    )

def get_pdn_medium(out_channels=384, padding=False):
    pad_mult = 1 if padding else 0
    return nn.Sequential(
        nn.Conv2d(in_channels=3, out_channels=256, kernel_size=4,
                  padding=3 * pad_mult),
        nn.ReLU(inplace=True),
        nn.AvgPool2d(kernel_size=2, stride=2, padding=1 * pad_mult),
        nn.Conv2d(in_channels=256, out_channels=512, kernel_size=4,
                  padding=3 * pad_mult),
        nn.ReLU(inplace=True),
        nn.AvgPool2d(kernel_size=2, stride=2, padding=1 * pad_mult),
        nn.Conv2d(in_channels=512, out_channels=512, kernel_size=1),
        nn.ReLU(inplace=True),
        nn.Conv2d(in_channels=512, out_channels=512, kernel_size=3,
                  padding=1 * pad_mult),
        nn.ReLU(inplace=True),
        nn.Conv2d(in_channels=512, out_channels=out_channels, kernel_size=4),
        nn.ReLU(inplace=True),
        nn.Conv2d(in_channels=out_channels, out_channels=out_channels,
                  kernel_size=1)
    )

class ImageFolderWithoutTarget(ImageFolder):
    def __getitem__(self, index):
        sample, target = super().__getitem__(index)
        return sample

class SingleFolderWithoutTarget(VisionDataset):
    def __init__(self, root, subdirectory, transform=None, target_transform=None):
        super(SingleFolderWithoutTarget, self).__init__(root, transform=transform, target_transform=target_transform)
        self.root = os.path.join(root, subdirectory)
        self.image_paths = [os.path.join(self.root, filename) for filename in os.listdir(self.root)]

    def __len__(self):
        return len(self.image_paths)

    def __getitem__(self, index):
        path = self.image_paths[index]
        sample = Image.open(path)
        if self.transform is not None:
            sample = self.transform(sample)
        return sample


class ImageFolderWithPath(ImageFolder):
    def __getitem__(self, index):
        path, target = self.samples[index]
        sample, target = super().__getitem__(index)
        return sample, target, path

def InfiniteDataloader(loader):
    iterator = iter(loader)
    while True:
        try:
            yield next(iterator)
        except StopIteration:
            iterator = iter(loader)

# F1 evalution code from https://github.com/caoyunkang/WinClip
def calculate_f1_max(gt, scores):
    precision, recall, thresholds = precision_recall_curve(gt, scores)
    a = 2 * precision * recall
    b = precision + recall
    f1s = np.divide(a, b, out=np.zeros_like(a), where=b != 0)
    index = np.argmax(f1s)
    max_f1 = f1s[index]
    threshold = thresholds[index]
    return max_f1, threshold