基于ConvNeXt的语义分割代码实现

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
import torch
import torch.nn as nn
import torch.nn.functional as F
from timm.models.layers import DropPath, trunc_normal_
class layer_Norm(nn.Module):
def __init__(self, normalized_shape, eps=1e-6, data_format='channels_last'):
super(layer_Norm, self).__init__()
self.weight =nn.Parameter(torch.ones(normalized_shape), requires_grad=True)
self.bias = nn.Parameter(torch.zeros(normalized_shape), requires_grad=True)
self.eps = eps
self.data_format = data_format
if self.data_format not in ['channels_last', 'channels_first']:
raise NotImplementedError
self.normalized_shape = (normalized_shape, )
def forward(self, x):
# [batch_size, height, weight, channel]
if self.data_format == 'channels_last':
return F.layer_norm(x, self.normalized_shape, self.weight, self.bias, self.eps)
elif self.data_format == 'channels_first':
mean = x.mean(1, keepdim=True)
var = (x - mean).pow(2).mean(1, keepdim=True)
x = (x - mean) / torch.sqrt(var + self.eps)
x = self.weight[:, None, None] * x + self.bias[:, None, None]
return x
class Block(nn.Module):
def __init__(self, dim, drop_path=0., layer_scale_init_value=1e-6):
super(Block, self).__init__()
self.dwconv = nn.Conv2d(dim, dim, kernel_size=7, padding=3, groups=dim)
self.norm = layer_Norm(dim, eps=1e-6)
self.pwconv1 = nn.Linear(dim, 4 * dim)
self.act = nn.GELU()
self.pwconv2 = nn.Linear(4 * dim, dim)
self.gamma = nn.Parameter(layer_scale_init_value * torch.ones((dim)), requires_grad=True) if layer_scale_init_value > 0 else None
self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
def forward(self, x):
input = x
x = self.dwconv(x)
x = x.permute(0, 2, 3, 1)
x = self.norm(x)
x = self.pwconv1(x)
x = self.act(x)
x = self.pwconv2(x)
if self.gamma is not None:
x = self.gamma * x
x = x.permute(0, 3, 1, 2)
x = input + self.drop_path(x)
return x
class ConvNeXt(nn.Module):
def __init__(self, in_channels=3, num_class=12, depths=[3, 3, 9, 3], dims=[96, 192, 384, 768], drop_path_rate=0., layer_scale_init_value=1e-6, head_init_scale=1.):
super(ConvNeXt, self).__init__()
self.downsample_layers = nn.ModuleList()
stem = nn.Sequential(
nn.Conv2d(in_channels, dims[0], kernel_size=4, stride=4),
layer_Norm(dims[0], eps=1e-6, data_format='channels_first')
)
self.downsample_layers.append(stem)
for i in range(3):
downsample_layer = nn.Sequential(
layer_Norm(dims[i], eps=1e-6, data_format='channels_first'),
nn.Conv2d(dims[i], dims[i+1], kernel_size=2, stride=2)
)
self.downsample_layers.append(downsample_layer)
self.stages = nn.ModuleList()
dp_rates = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))]
cur = 0
for i in range(4):
stage = nn.Sequential(
*[Block(dim=dims[i], drop_path=dp_rates[cur + j], layer_scale_init_value=layer_scale_init_value) for j in range(depths[i])]
)
self.stages.append(stage)
cur += depths[i]
# self.norm = nn.LayerNorm(dims[-1], eps=1e-6)
self.norm = layer_Norm(dims[-1], eps=1e-6, data_format='channels_first')
self.head = nn.Linear(dims[-1], num_class)
self.apply(self._init_weights)
self.x4_12 = nn.Conv2d(768, num_class, kernel_size=1)
self.norm_x4_12 = layer_Norm(num_class, eps=1e-6, data_format='channels_first')
self.act = nn.GELU()
self.deconv_x4 = nn.ConvTranspose2d(num_class, num_class, 4, 2, 1)
self.norm_x3 = layer_Norm(384, eps=1e-6, data_format='channels_first')
self.x3_12 = nn.Conv2d(384, num_class, kernel_size=1)
self.deconv_x3 = nn.ConvTranspose2d(num_class, num_class, 4, 2, 1)
self.norm_x2 = layer_Norm(192, eps=1e-6, data_format='channels_first')
self.x2_12 = nn.Conv2d(192, num_class, kernel_size=1)
self.deconv_x2 = nn.ConvTranspose2d(num_class, num_class, 4, 2, 1)
self.norm_x1 = layer_Norm(96, eps=1e-6, data_format='channels_first')
self.x1_12 = nn.Conv2d(96, num_class, kernel_size=1)
self.upsample = nn.ConvTranspose2d(num_class, num_class, 8, 4, 2, bias=False)
def _init_weights(self, m):
if isinstance(m, (nn.Conv2d, nn.Linear)):
trunc_normal_(m.weight, std=.02)
nn.init.constant_(m.bias, 0)
def forward_features(self, x):
x = self.downsample_layers[0](x)
x = self.stages[0](x)
x1 = x
x = self.downsample_layers[1](x)
x = self.stages[1](x)
x2 = x
x = self.downsample_layers[2](x)
x = self.stages[2](x)
x3 = x
x = self.downsample_layers[3](x)
x = self.stages[3](x)
x = self.norm(x)
return x, x3, x2, x1
def forward(self, x):
x, x3, x2, x1 = self.forward_features(x)
x4_12 = self.act(self.norm_x4_12(self.x4_12(x)))
x4_x3 = self.act(self.norm_x4_12(self.deconv_x4(x4_12)))
x3_norm = self.norm_x3(x3)
x3_12 = self.act(self.norm_x4_12(self.x3_12(x3_norm)))
x3_x4 = x4_x3 + x3_12
x43 = self.deconv_x3(x3_x4)
x2_norm = self.norm_x2(x2)
x2_12 = self.act(self.norm_x4_12(self.x2_12(x2_norm)))
x2_x3 = x2_12 + x43
x32 = self.deconv_x2(x2_x3)
x1_norm = self.norm_x1(x1)
x1_12 = self.act(self.norm_x4_12(self.x1_12(x1_norm)))
x2_x1 = x32 + x1_12
x = self.upsample(x2_x1)
# print(x.shape)
# print(x4.shape)
# print(x3.shape)
# print(x2.shape)
# print(x1.shape)
return x
if __name__ == '__main__':
rgb = torch.randn([1, 3, 224, 224])
net = ConvNeXt()
out = net(rgb)
print(out.shape)