HFC-Net/HFC-Net.py at main · 7dra/HFC-Net · 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
import torch
import torch.nn as nn
import torch.nn.functional as F


class FR(nn.Module):
    def __init__(self, channels):
        super(FR, self).__init__()
        self.mask_generator = nn.Sequential(
            nn.Conv2d(channels, channels // 8, kernel_size=3, padding=1),
            nn.ReLU(),
            nn.Conv2d(channels // 8, channels, kernel_size=1),
            nn.Sigmoid()
        )
        self.reconstructor = nn.Sequential(
            nn.Conv2d(channels, channels, kernel_size=3, padding=1, groups=channels),
            nn.BatchNorm2d(channels),
            nn.ReLU()
        )

    def forward(self, x):

        mask = self.mask_generator(x)

        recon = self.reconstructor(x)

        masked_features = x * mask

        reconstructed_features = self.reconstructor(masked_features)


        return x + reconstructed_features


class GCA(nn.Module):
    def __init__(self, channels):
        super(GCA, self).__init__()
        self.norm = nn.LayerNorm(channels)
        self.attention = nn.MultiheadAttention(channels, num_heads=4)
        self.ffn = nn.Sequential(
            nn.Conv2d(channels, channels * 2, kernel_size=1),
            nn.ReLU(),
            nn.Conv2d(channels * 2, channels, kernel_size=1)
        )

    def forward(self, x):
        b, c, h, w = x.size()
        x_flat = x.view(b, c, -1).permute(2, 0, 1)  # (H*W, B, C)
        x_flat = self.norm(x_flat)

        attn_output, _ = self.attention(x_flat, x_flat, x_flat)
        attn_output = attn_output.permute(1, 2, 0).view(b, c, h, w)

        out = x + attn_output

        out = out + self.ffn(out)
        return out


class ConvBlock(nn.Module):
    def __init__(self, in_channels, out_channels):
        super(ConvBlock, self).__init__()
        self.conv = nn.Sequential(
            nn.Conv2d(in_channels, out_channels, kernel_size=3, padding=1),
            nn.BatchNorm2d(out_channels),
            nn.ReLU(),
            nn.Conv2d(out_channels, out_channels, kernel_size=3, padding=1),
            nn.BatchNorm2d(out_channels),
            nn.ReLU()
        )

    def forward(self, x):
        return self.conv(x)

class HFC(nn.Module):
    def __init__(self, num_classes):
        super(HFC, self).__init__()
        # encoder
        self.encoder1 = ConvBlock(3, 64)
        self.encoder2 = ConvBlock(64, 128)
        self.encoder3 = ConvBlock(128, 256)
        self.encoder4 = ConvBlock(256, 512)


        self.fr = FR(512)
        self.gca = GCA(512)

        # decoder
        self.upconv3 = nn.ConvTranspose2d(512, 256, kernel_size=2, stride=2)
        self.decoder3 = ConvBlock(512, 256)

        self.upconv2 = nn.ConvTranspose2d(256, 128, kernel_size=2, stride=2)
        self.decoder2 = ConvBlock(256, 128)

        self.upconv1 = nn.ConvTranspose2d(128, 64, kernel_size=2, stride=2)
        self.decoder1 = ConvBlock(128, 64)

        self.final = nn.Conv2d(64, num_classes, kernel_size=1)

    def forward(self, x):

        e1 = self.encoder1(x)

        e2 = self.encoder2(F.max_pool2d(e1, 2))

        e3 = self.encoder3(F.max_pool2d(e2, 2))

        e4 = self.encoder4(F.max_pool2d(e3, 2))


        e4 = self.fr(e4)

        e4 = self.gca(e4)

        d3 = self.upconv3(e4)

        d3 = torch.cat([d3, e3], dim=1)

        d3 = self.decoder3(d3)


        d2 = self.upconv2(d3)

        d2 = torch.cat([d2, e2], dim=1)

        d2 = self.decoder2(d2)

        d1 = self.upconv1(d2)
        d1 = torch.cat([d1, e1], dim=1)

        d1 = self.decoder1(d1)
        out = self.final(d1)
        return out

x = torch.randn(2, 3, 224, 224)
net = HFC(3)
out = net(x)
print(out.shape)