-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathmodel_coatsmall_daformer.py
165 lines (130 loc) · 4.7 KB
/
model_coatsmall_daformer.py
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
from kaggle_hubmap_kv3 import *
from daformer import *
from coat_small import *
import copy
#################################################################
def criterion_aux_loss(logit, mask):
mask = F.interpolate(mask,size=logit.shape[-2:], mode='nearest')
loss = F.binary_cross_entropy_with_logits(logit,mask)
return loss
class RGB(nn.Module):
IMAGE_RGB_MEAN = [0.485, 0.456, 0.406] # [0.5, 0.5, 0.5]
IMAGE_RGB_STD = [0.229, 0.224, 0.225] # [0.5, 0.5, 0.5]
def __init__(self, ):
super(RGB, self).__init__()
self.register_buffer('mean', torch.zeros(1, 3, 1, 1))
self.register_buffer('std', torch.ones(1, 3, 1, 1))
self.mean.data = torch.FloatTensor(self.IMAGE_RGB_MEAN).view(self.mean.shape)
self.std.data = torch.FloatTensor(self.IMAGE_RGB_STD).view(self.std.shape)
def forward(self, x):
x = (x - self.mean) / self.std
return x
class Net(nn.Module):
def __init__(self,
encoder=coat_small,
decoder=daformer_conv1x1,
encoder_cfg={},
decoder_cfg={},
encoder_ckpt=None,
decoder_ckpt=None
):
super(Net, self).__init__()
decoder_dim = decoder_cfg.get('decoder_dim', 320)
# ----
self.rgb = RGB()
self.encoder = encoder(
#drop_path_rate=0.3,
)
if encoder_ckpt is not None:
checkpoint = torch.load(encoder_ckpt, map_location=lambda storage, loc: storage)
self.encoder.load_state_dict(checkpoint['model'],strict=False)
self.encoders = torch.nn.ModuleList()
# for _ in range(2):
self.encoders.append(self.encoder)
self.encoders.append(copy.deepcopy(self.encoders[0]))
encoder_dim = self.encoder.embed_dims
# [64, 128, 320, 512]
self.decoder = decoder(
encoder_dim=encoder_dim,
decoder_dim=decoder_dim,
)
# self.decoders = torch.nn.ModuleList()
# self.decoders.append(self.decoder)
# self.decoders.append(copy.deepcopy(self.decoders[0]))
# self.decoders = torch.nn.ModuleList()
# for _ in range(5):
# self.decoders.append(copy.deepcopy(self.decoder))
self.logit = nn.Sequential(
nn.Conv2d(decoder_dim, 1, kernel_size=1),
)
self.output_type = ['inference', 'loss']
self.aux = nn.ModuleList([
nn.Conv2d(decoder_dim, 1, kernel_size=1, padding=0) for i in range(4)
])
def forward(self, batch):
# import ipdb;ipdb.set_trace()
# mask = batch['mask']
organs = batch['organ']
x = batch['image']
x = self.rgb(x)
B, C, H, W = x.shape
# import ipdb;ipdb.set_trace()
# if self.training:
# # encoder = []
# # for i in range(B):
# encoder = self.encoders[organs.item()](x)
# # import ipdb;ipdb.set_trace()
# # encoder = torch.concat(encoder)
# else:
encoder = self.encoders[organs[0].item() // 4](x)
#print([f.shape for f in encoder])
# import ipdb;ipdb.set_trace()
last, decoder = self.decoder(encoder)
# last, decoder = self.decoders[organs[0].item()](encoder)
logit = self.logit(last)
# import ipdb;ipdb.set_trace()
# print(logit.shape)
logit2 = F.interpolate(logit, size=None, scale_factor=4, mode='bilinear', align_corners=False)
# mask = F.interpolate(mask, size=None, scale_factor=1/4, mode='bilinear', align_corners=False)
output = {}
# probability_from_logit = torch.sigmoid(logit)
# output['probability'] = probability_from_logit
# import ipdb;ipdb.set_trace()
if 'loss' in self.output_type:
# import ipdb;ipdb.set_trace()
mask = batch['mask']
mask = F.interpolate(mask, size=None, scale_factor=1/4, mode='bilinear', align_corners=False)
# output['bce_loss'] = F.binary_cross_entropy_with_logits(logit, batch['mask'])
output['bce_loss'] = F.binary_cross_entropy_with_logits(logit, mask)
for i in range(4):
output['aux%d_loss'%i] = criterion_aux_loss(self.aux[i](decoder[i]),batch['mask'])
if 'inference' in self.output_type:
output['probability'] = torch.sigmoid(logit2)
return output
def run_check_net():
batch_size = 2
image_size = 800
# ---
batch = {
'image': torch.from_numpy(np.random.uniform(-1, 1, (batch_size, 3, image_size, image_size))).float(),
'mask': torch.from_numpy(np.random.choice(2, (batch_size, 1, image_size, image_size))).float(),
'organ': torch.from_numpy(np.random.choice(5, (batch_size, 1))).long(),
}
batch = {k: v.cuda() for k, v in batch.items()}
net = Net().cuda()
with torch.no_grad():
with torch.cuda.amp.autocast(enabled=True):
output = net(batch)
print('batch')
for k, v in batch.items():
print('%32s :' % k, v.shape)
print('output')
for k, v in output.items():
if 'loss' not in k:
print('%32s :' % k, v.shape)
for k, v in output.items():
if 'loss' in k:
print('%32s :' % k, v.item())
# main #################################################################
if __name__ == '__main__':
run_check_net()