File size: 4,954 Bytes
e99a83c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
import torch
import torch.nn as nn
import torch.nn.functional as F


class ConvBNReLU(nn.Module):
    def __init__(self, in_channels, out_channels, kernel_size=3, padding=1):
        super().__init__()

        self.block = nn.Sequential(
            nn.Conv2d(in_channels, out_channels, kernel_size=kernel_size, padding=padding, bias=False),
            nn.BatchNorm2d(out_channels),
            nn.ReLU(inplace=True),
        )

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


class ResidualBlock(nn.Module):
    """
    Basic residual block for ResUNet.

    If in_channels != out_channels, the shortcut uses a 1x1 conv.
    """

    def __init__(self, in_channels, out_channels):
        super().__init__()

        self.conv1 = ConvBNReLU(in_channels, out_channels)
        self.conv2 = nn.Sequential(
            nn.Conv2d(out_channels, out_channels, kernel_size=3, padding=1, bias=False),
            nn.BatchNorm2d(out_channels),
        )

        if in_channels != out_channels:
            self.shortcut = nn.Sequential(
                nn.Conv2d(in_channels, out_channels, kernel_size=1, bias=False),
                nn.BatchNorm2d(out_channels),
            )
        else:
            self.shortcut = nn.Identity()

        self.relu = nn.ReLU(inplace=True)

    def forward(self, x):
        residual = self.shortcut(x)

        x = self.conv1(x)
        x = self.conv2(x)

        x = x + residual
        x = self.relu(x)

        return x


class EncoderBlock(nn.Module):
    def __init__(self, in_channels, out_channels):
        super().__init__()

        self.res_block = ResidualBlock(in_channels, out_channels)
        self.pool = nn.MaxPool2d(kernel_size=2, stride=2)

    def forward(self, x):
        skip = self.res_block(x)
        pooled = self.pool(skip)
        return skip, pooled


class DecoderBlock(nn.Module):
    def __init__(self, in_channels, skip_channels, out_channels):
        super().__init__()

        self.up = nn.ConvTranspose2d(
            in_channels,
            out_channels,
            kernel_size=2,
            stride=2,
        )

        self.res_block = ResidualBlock(
            out_channels + skip_channels,
            out_channels,
        )

    def forward(self, x, skip):
        x = self.up(x)

        # Handles odd image sizes, though 512/1024 should already match.
        if x.shape[-2:] != skip.shape[-2:]:
            x = F.interpolate(
                x,
                size=skip.shape[-2:],
                mode="bilinear",
                align_corners=False,
            )

        x = torch.cat([x, skip], dim=1)
        x = self.res_block(x)

        return x


class ResUNet(nn.Module):
    """
    ResUNet for binary or multi-class retinal segmentation.

    Output:
        Raw logits of shape [B, num_classes, H, W]

    For vessel segmentation:
        num_classes=1
        loss=BCEWithLogits/Dice/Tversky/etc.
    """

    def __init__(
        self,
        in_channels=3,
        num_classes=1,
        base_channels=32,
        dropout=0.0,
    ):
        super().__init__()

        c1 = base_channels
        c2 = base_channels * 2
        c3 = base_channels * 4
        c4 = base_channels * 8
        c5 = base_channels * 16

        self.enc1 = EncoderBlock(in_channels, c1)
        self.enc2 = EncoderBlock(c1, c2)
        self.enc3 = EncoderBlock(c2, c3)
        self.enc4 = EncoderBlock(c3, c4)

        self.bottleneck = nn.Sequential(
            ResidualBlock(c4, c5),
            nn.Dropout2d(dropout),
        )

        self.dec4 = DecoderBlock(c5, c4, c4)
        self.dec3 = DecoderBlock(c4, c3, c3)
        self.dec2 = DecoderBlock(c3, c2, c2)
        self.dec1 = DecoderBlock(c2, c1, c1)

        self.out_conv = nn.Conv2d(c1, num_classes, kernel_size=1)

    def forward(self, x):
        s1, x = self.enc1(x)
        s2, x = self.enc2(x)
        s3, x = self.enc3(x)
        s4, x = self.enc4(x)

        x = self.bottleneck(x)

        x = self.dec4(x, s4)
        x = self.dec3(x, s3)
        x = self.dec2(x, s2)
        x = self.dec1(x, s1)

        logits = self.out_conv(x)

        return logits


def build_resunet(
    in_channels=3,
    num_classes=1,
    base_channels=32,
    dropout=0.0,
):
    return ResUNet(
        in_channels=in_channels,
        num_classes=num_classes,
        base_channels=base_channels,
        dropout=dropout,
    )


if __name__ == "__main__":
    # Smoke test:
    # python models/unet.py

    device = "cuda" if torch.cuda.is_available() else "cpu"

    model = build_resunet(
        in_channels=3,
        num_classes=1,
        base_channels=32,
        dropout=0.0,
    ).to(device)

    x = torch.randn(2, 3, 512, 512).to(device)

    with torch.no_grad():
        y = model(x)

    print("Input shape:", x.shape)
    print("Output shape:", y.shape)
    print("Output min/max:", y.min().item(), y.max().item())

    assert y.shape == (2, 1, 512, 512)

    print("Smoke test passed.")