deep-transfer/encoder_decoder_factory.py at master · pietrocarbo/deep-transfer · 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
import torch
import torch.nn as nn
from models.autoencoder_vgg19.vgg19_1 import vgg_normalised_conv1_1, feature_invertor_conv1_1
from models.autoencoder_vgg19.vgg19_2 import vgg_normalised_conv2_1, feature_invertor_conv2_1
from models.autoencoder_vgg19.vgg19_3 import vgg_normalised_conv3_1, feature_invertor_conv3_1
from models.autoencoder_vgg19.vgg19_4 import vgg_normalised_conv4_1, feature_invertor_conv4_1
from models.autoencoder_vgg19.vgg19_5 import vgg_normalised_conv5_1, feature_invertor_conv5_1

class Encoder(nn.Module):

    def __init__(self, depth):
        super(Encoder, self).__init__()

        assert(type(depth).__name__ == 'int' and 1 <= depth <= 5)
        self.depth = depth

        if depth == 1:
            self.model = vgg_normalised_conv1_1.vgg_normalised_conv1_1
            self.model.load_state_dict(torch.load("models/autoencoder_vgg19/vgg19_1/vgg_normalised_conv1_1.pth"))
        elif depth == 2:
            self.model = vgg_normalised_conv2_1.vgg_normalised_conv2_1
            self.model.load_state_dict(torch.load("models/autoencoder_vgg19/vgg19_2/vgg_normalised_conv2_1.pth"))
        elif depth == 3:
            self.model = vgg_normalised_conv3_1.vgg_normalised_conv3_1
            self.model.load_state_dict(torch.load("models/autoencoder_vgg19/vgg19_3/vgg_normalised_conv3_1.pth"))
        elif depth == 4:
            self.model = vgg_normalised_conv4_1.vgg_normalised_conv4_1
            self.model.load_state_dict(torch.load("models/autoencoder_vgg19/vgg19_4/vgg_normalised_conv4_1.pth"))
        elif depth == 5:
            self.model = vgg_normalised_conv5_1.vgg_normalised_conv5_1
            self.model.load_state_dict(torch.load("models/autoencoder_vgg19/vgg19_5/vgg_normalised_conv5_1.pth"))


    def forward(self, x):
        out = self.model(x)
        return out


class Decoder(nn.Module):
    def __init__(self, depth):
        super(Decoder, self).__init__()

        assert (type(depth).__name__ == 'int' and 1 <= depth <= 5)
        self.depth = depth

        if depth == 1:
            self.model = feature_invertor_conv1_1.feature_invertor_conv1_1
            self.model.load_state_dict(torch.load("models/autoencoder_vgg19/vgg19_1/feature_invertor_conv1_1.pth"))
        elif depth == 2:
            self.model = feature_invertor_conv2_1.feature_invertor_conv2_1
            self.model.load_state_dict(torch.load("models/autoencoder_vgg19/vgg19_2/feature_invertor_conv2_1.pth"))
        elif depth == 3:
            self.model = feature_invertor_conv3_1.feature_invertor_conv3_1
            self.model.load_state_dict(torch.load("models/autoencoder_vgg19/vgg19_3/feature_invertor_conv3_1.pth"))
        elif depth == 4:
            self.model = feature_invertor_conv4_1.feature_invertor_conv4_1
            self.model.load_state_dict(torch.load("models/autoencoder_vgg19/vgg19_4/feature_invertor_conv4_1.pth"))
        elif depth == 5:
            self.model = feature_invertor_conv5_1.feature_invertor_conv5_1
            self.model.load_state_dict(torch.load("models/autoencoder_vgg19/vgg19_5/feature_invertor_conv5_1.pth"))

    def forward(self, x):
        out = self.model(x)
        return out