DiffProxy/vgg.py at main · yiwei-hu/DiffProxy · 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
160
161
162
163
164
import os
import torch
import torch.nn as nn
import torch.nn.functional as F

vgg_path = 'pretrained/vgg_conv.pt'

# vgg definition that conveniently let's you grab the outputs from any layer
class VGG(nn.Module):
    def __init__(self, pool='max'):
        super(VGG, self).__init__()
        # vgg modules
        self.conv1_1 = nn.Conv2d(3, 64, kernel_size=3, padding=1)
        self.conv1_2 = nn.Conv2d(64, 64, kernel_size=3, padding=1)
        self.conv2_1 = nn.Conv2d(64, 128, kernel_size=3, padding=1)
        self.conv2_2 = nn.Conv2d(128, 128, kernel_size=3, padding=1)
        self.conv3_1 = nn.Conv2d(128, 256, kernel_size=3, padding=1)
        self.conv3_2 = nn.Conv2d(256, 256, kernel_size=3, padding=1)
        self.conv3_3 = nn.Conv2d(256, 256, kernel_size=3, padding=1)
        self.conv3_4 = nn.Conv2d(256, 256, kernel_size=3, padding=1)
        self.conv4_1 = nn.Conv2d(256, 512, kernel_size=3, padding=1)
        self.conv4_2 = nn.Conv2d(512, 512, kernel_size=3, padding=1)
        self.conv4_3 = nn.Conv2d(512, 512, kernel_size=3, padding=1)
        self.conv4_4 = nn.Conv2d(512, 512, kernel_size=3, padding=1)
        self.conv5_1 = nn.Conv2d(512, 512, kernel_size=3, padding=1)
        self.conv5_2 = nn.Conv2d(512, 512, kernel_size=3, padding=1)
        self.conv5_3 = nn.Conv2d(512, 512, kernel_size=3, padding=1)
        self.conv5_4 = nn.Conv2d(512, 512, kernel_size=3, padding=1)
        if pool == 'max':
            self.pool1 = nn.MaxPool2d(kernel_size=2, stride=2)
            self.pool2 = nn.MaxPool2d(kernel_size=2, stride=2)
            self.pool3 = nn.MaxPool2d(kernel_size=2, stride=2)
            self.pool4 = nn.MaxPool2d(kernel_size=2, stride=2)
            self.pool5 = nn.MaxPool2d(kernel_size=2, stride=2)
        elif pool == 'avg':
            self.pool1 = nn.AvgPool2d(kernel_size=2, stride=2)
            self.pool2 = nn.AvgPool2d(kernel_size=2, stride=2)
            self.pool3 = nn.AvgPool2d(kernel_size=2, stride=2)
            self.pool4 = nn.AvgPool2d(kernel_size=2, stride=2)
            self.pool5 = nn.AvgPool2d(kernel_size=2, stride=2)

        self.normalizer = VGG19Normalizer()

    def forward(self, x, out_keys):
        out = {}
        out['r11'] = F.relu(self.conv1_1(x))
        out['r12'] = F.relu(self.conv1_2(out['r11']))
        out['p1'] = self.pool1(out['r12'])
        out['r21'] = F.relu(self.conv2_1(out['p1']))
        out['r22'] = F.relu(self.conv2_2(out['r21']))
        out['p2'] = self.pool2(out['r22'])
        out['r31'] = F.relu(self.conv3_1(out['p2']))
        out['r32'] = F.relu(self.conv3_2(out['r31']))
        out['r33'] = F.relu(self.conv3_3(out['r32']))
        out['r34'] = F.relu(self.conv3_4(out['r33']))
        out['p3'] = self.pool3(out['r34'])
        out['r41'] = F.relu(self.conv4_1(out['p3']))
        out['r42'] = F.relu(self.conv4_2(out['r41']))
        out['r43'] = F.relu(self.conv4_3(out['r42']))
        out['r44'] = F.relu(self.conv4_4(out['r43']))
        out['p4'] = self.pool4(out['r44'])
        out['r51'] = F.relu(self.conv5_1(out['p4']))
        out['r52'] = F.relu(self.conv5_2(out['r51']))
        out['r53'] = F.relu(self.conv5_3(out['r52']))
        out['r54'] = F.relu(self.conv5_4(out['r53']))
        out['p5'] = self.pool5(out['r54'])
        return [out[key] for key in out_keys]

    def extract_features(self, im, feature_layers, style_layers, detach_features=False, detach_styles=False):
        if im.shape[1] == 1:
            im = torch.cat((im, im, im), dim=1)

        x = self.normalizer.normalize(im)

        out = {}
        out['r11'] = F.relu(self.conv1_1(x))
        out['r12'] = F.relu(self.conv1_2(out['r11']))
        out['p1'] = self.pool1(out['r12'])
        out['r21'] = F.relu(self.conv2_1(out['p1']))
        out['r22'] = F.relu(self.conv2_2(out['r21']))
        out['p2'] = self.pool2(out['r22'])
        out['r31'] = F.relu(self.conv3_1(out['p2']))
        out['r32'] = F.relu(self.conv3_2(out['r31']))
        out['r33'] = F.relu(self.conv3_3(out['r32']))
        out['r34'] = F.relu(self.conv3_4(out['r33']))
        out['p3'] = self.pool3(out['r34'])
        out['r41'] = F.relu(self.conv4_1(out['p3']))
        out['r42'] = F.relu(self.conv4_2(out['r41']))
        out['r43'] = F.relu(self.conv4_3(out['r42']))
        out['r44'] = F.relu(self.conv4_4(out['r43']))
        out['p4'] = self.pool4(out['r44'])
        out['r51'] = F.relu(self.conv5_1(out['p4']))
        out['r52'] = F.relu(self.conv5_2(out['r51']))
        out['r53'] = F.relu(self.conv5_3(out['r52']))
        out['r54'] = F.relu(self.conv5_4(out['r53']))
        out['p5'] = self.pool5(out['r54'])
        if not detach_features:
            extracted_features = [out[layer] for layer in feature_layers]
        else:
            extracted_features = [out[layer].detach() for layer in feature_layers]
        if not detach_styles:
            extracted_styles = [gram(out[layer]) for layer in style_layers]
        else:
            extracted_styles = [gram(out[layer]).detach() for layer in style_layers]
        return extracted_features, extracted_styles


# get network
def get_vgg19():
    vgg = VGG()
    vgg.load_state_dict(torch.load(vgg_path))
    for param in vgg.parameters():
        param.requires_grad = False
    if torch.cuda.is_available():
        vgg.cuda()
    return vgg


def gram(x):
    b, c, h, w = x.shape
    F = x.view(b, c, h*w)
    G = torch.bmm(F, F.transpose(1,2))
    G.div_(h*w)
    return G


class WeightedLoss(nn.Module):
    def __init__(self, weights, metric='l2'):
        super(WeightedLoss, self).__init__()

        self.weights = weights
        if metric == 'l2':
            self.criterion = nn.MSELoss().cuda()
        elif metric == 'l1':
            self.criterion = nn.L1Loss().cuda()
        else:
            raise NotImplementedError('Unknown metric {}'.format(metric))

    def forward(self, x, y):
        loss = torch.tensor(0.0, requires_grad=True).cuda()
        for w, x_, y_ in zip(self.weights, x, y):
            loss = loss + w * self.criterion(x_, y_)
        return loss


class VGG19Normalizer(nn.Module):
    def __init__(self):
        super(VGG19Normalizer,self).__init__()
        imagenet_mean = [0.40760392, 0.45795686, 0.48501961]
        self.mean = torch.as_tensor(imagenet_mean)[None, :, None, None].cuda()

    def normalize(self, image):
        image = image * 0.5 + 0.5 # (-1, 1) to (0, 1)
        image = image[:, [2, 1, 0], :, :]  # turn to BGR
        image = image - self.mean  # subtract imagenet mean
        image = image * 255.0
        return image

    def denormalize(self, image):
        image = image / 255.0
        image = image + self.mean  # add imagenet mean
        image = image[:, [2, 1, 0], :, :]  # turn to RGB
        image = image * 2.0 - 1.0 # (0, 1) to (-1, 1)
        return image