File size: 6,033 Bytes
8ed2f16
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import torch
import torch.nn as nn
import torch.nn.functional as F
import torchvision
from lib.models.networks.encoder import VGGEncoder
# from util import util
from lib.models.networks.sync_batchnorm import SynchronizedBatchNorm2d
import torch.nn.utils.spectral_norm as spectral_norm

def copy_state_dict(state_dict, model, strip=None, replace=None):
    tgt_state = model.state_dict()
    copied_names = set()
    for name, param in state_dict.items():
        if strip is not None and replace is None and name.startswith(strip):
            name = name[len(strip):]
        if strip is not None and replace is not None:
            name = name.replace(strip, replace)
        if name not in tgt_state:
            continue
        if isinstance(param, torch.nn.Parameter):
            param = param.data
        if param.size() != tgt_state[name].size():
            print('mismatch:', name, param.size(), tgt_state[name].size())
            continue
        tgt_state[name].copy_(param)
        copied_names.add(name)

    missing = set(tgt_state.keys()) - copied_names
    if len(missing) > 0:
        print("missing keys in state_dict:", missing)

# VGG architecter, used for the perceptual loss using a pretrained VGG network
class VGG19(torch.nn.Module):
    def __init__(self, requires_grad=False):
        super(VGG19, self).__init__()
        vgg_pretrained_features = torchvision.models.vgg19(pretrained=True).features
        self.slice1 = torch.nn.Sequential()
        self.slice2 = torch.nn.Sequential()
        self.slice3 = torch.nn.Sequential()
        self.slice4 = torch.nn.Sequential()
        self.slice5 = torch.nn.Sequential()
        for x in range(2):
            self.slice1.add_module(str(x), vgg_pretrained_features[x])
        for x in range(2, 7):
            self.slice2.add_module(str(x), vgg_pretrained_features[x])
        for x in range(7, 12):
            self.slice3.add_module(str(x), vgg_pretrained_features[x])
        for x in range(12, 21):
            self.slice4.add_module(str(x), vgg_pretrained_features[x])
        for x in range(21, 30):
            self.slice5.add_module(str(x), vgg_pretrained_features[x])
        if not requires_grad:
            for param in self.parameters():
                param.requires_grad = False

    def forward(self, X):
        h_relu1 = self.slice1(X)
        h_relu2 = self.slice2(h_relu1)
        h_relu3 = self.slice3(h_relu2)
        h_relu4 = self.slice4(h_relu3)
        h_relu5 = self.slice5(h_relu4)
        out = [h_relu1, h_relu2, h_relu3, h_relu4, h_relu5]
        return out


class VGGFace19(torch.nn.Module):
    def __init__(self, opt, load_path="", requires_grad=False):
        super(VGGFace19, self).__init__()
        self.model = VGGEncoder(opt)
        self.opt = opt
        ckpt = torch.load(load_path)
        print("=> loading checkpoint '{}'".format(load_path))
        copy_state_dict(ckpt, self.model.model)
        vgg_pretrained_features = self.model.model.features
        len_features = len(self.model.model.features)
        self.slice1 = torch.nn.Sequential()
        self.slice2 = torch.nn.Sequential()
        self.slice3 = torch.nn.Sequential()
        self.slice4 = torch.nn.Sequential()
        self.slice5 = torch.nn.Sequential()
        self.slice6 = torch.nn.Sequential()

        for x in range(2):
            self.slice1.add_module(str(x), vgg_pretrained_features[x])
        for x in range(2, 7):
            self.slice2.add_module(str(x), vgg_pretrained_features[x])
        for x in range(7, 12):
            self.slice3.add_module(str(x), vgg_pretrained_features[x])
        for x in range(12, 21):
            self.slice4.add_module(str(x), vgg_pretrained_features[x])
        for x in range(21, 30):
            self.slice5.add_module(str(x), vgg_pretrained_features[x])
        for x in range(30, len_features):
            self.slice6.add_module(str(x), vgg_pretrained_features[x])
        if not requires_grad:
            for param in self.parameters():
                param.requires_grad = False

    def forward(self, X):
        X = X.view(-1, self.opt.model.output_nc, self.opt.data.img_size, self.opt.data.img_size)
        h_relu1 = self.slice1(X)
        h_relu2 = self.slice2(h_relu1)
        h_relu3 = self.slice3(h_relu2)
        h_relu4 = self.slice4(h_relu3)
        h_relu5 = self.slice5(h_relu4)
        h_relu6 = self.slice6(h_relu5)
        out = [h_relu3, h_relu4, h_relu5, h_relu6, h_relu6]
        return out


# Returns a function that creates a normalization function
# that does not condition on semantic map
def get_nonspade_norm_layer(opt, norm_type='instance'):
    # helper function to get # output channels of the previous layer
    def get_out_channel(layer):
        if hasattr(layer, 'out_channels'):
            return getattr(layer, 'out_channels')
        return layer.weight.size(0)

    # this function will be returned
    def add_norm_layer(layer):
        nonlocal norm_type
        if norm_type.startswith('spectral'):
            layer = spectral_norm(layer)
            subnorm_type = norm_type[len('spectral'):]
        else:
            subnorm_type = norm_type

        if subnorm_type == 'none' or len(subnorm_type) == 0:
            return layer

        # remove bias in the previous layer, which is meaningless
        # since it has no effect after normalization
        if getattr(layer, 'bias', None) is not None:
            delattr(layer, 'bias')
            layer.register_parameter('bias', None)

        if subnorm_type == 'batch':
            norm_layer = nn.BatchNorm2d(get_out_channel(layer), affine=True)
        elif subnorm_type == 'syncbatch':
            norm_layer = SynchronizedBatchNorm2d(get_out_channel(layer), affine=True)
        elif subnorm_type == 'instance':
            norm_layer = nn.InstanceNorm2d(get_out_channel(layer), affine=False)
        else:
            raise ValueError('normalization layer %s is not recognized' % subnorm_type)

        return nn.Sequential(layer, norm_layer)

    return add_norm_layer