-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathmodel.py
113 lines (94 loc) · 4.63 KB
/
model.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
from __future__ import print_function
import torch
import torch.nn as nn
import torch.nn.parallel
import torch.utils.data
from torch.autograd import Variable
import numpy as np
import torch.nn.functional as F
class Triplet(nn.Module):
def __init__(self, drop_prob=0.):
super(Triplet, self).__init__()
self.drop_prob = drop_prob
self.conv1 = nn.Conv2d(in_channels=2, out_channels=64, kernel_size=(2,3), stride=(1,1), bias=True)
self.bn1 = nn.BatchNorm2d(64)
self.relu1 = nn.ReLU(inplace=True)
# self.relu1 = nn.Sigmoid()
self.dropout1 = nn.Dropout2d(.1)
self.pool1 = nn.AvgPool2d((1,3))
self.conv2 = nn.Conv2d(in_channels=64, out_channels=128, kernel_size=(1,3), stride=(1,1), bias=True)
self.bn2 = nn.BatchNorm2d(128)
self.relu2 = nn.ReLU(inplace=True)
# self.relu2 = nn.Sigmoid()
self.dropout2 = nn.Dropout2d(.1)
self.pool2 = nn.AvgPool2d((1,3))
self.conv3 = nn.Conv2d(in_channels=128, out_channels=128, kernel_size=(1,3), stride=(1,1), bias=True)
self.bn3 = nn.BatchNorm2d(128)
self.relu3 = nn.ReLU(inplace=True)
# self.relu3 = nn.Sigmoid()
self.dropout3 = nn.Dropout2d(.1)
self.pool3 = nn.AvgPool2d((1,3))
self.conv4 = nn.Conv2d(in_channels=128, out_channels=64, kernel_size=(1,3), stride=(1,1), bias=False)
self.bn4 = nn.BatchNorm2d(64)
self.relu4 = nn.ReLU(inplace=True)
self.dropout4 = nn.Dropout2d(.1)
self.pool4 = nn.AvgPool2d((1,3))
self.conv5 = nn.Conv2d(in_channels=64, out_channels=64, kernel_size=(1,3), stride=(1,1), bias=False)
self.bn5 = nn.BatchNorm2d(64)
self.relu5 = nn.ReLU(inplace=True)
self.dropout5 = nn.Dropout2d(.1)
self.pool5 = nn.AvgPool2d((1,3))
# self.conv6 = nn.Conv2d(in_channels=256, out_channels=128, kernel_size=(1,3), stride=(1,1), bias=False)
# self.bn6 = nn.BatchNorm2d(128)
# self.relu6 = nn.ReLU(inplace=True)
# self.dropout6 = nn.Dropout2d(drop_prob)
# self.maxpool6 = nn.MaxPool2d((1,3))
# self.conv7 = nn.Conv2d(in_channels=128, out_channels=64, kernel_size=(1,3), stride=(1,1), bias=False)
# self.bn7 = nn.BatchNorm2d(64)
# self.relu7 = nn.ReLU(inplace=True)
# self.dropout7 = nn.Dropout2d(drop_prob)
# self.maxpool7 = nn.MaxPool2d((1,3))
self.fc1 = nn.Linear(512, 64)
# self.fc1 = nn.Linear(58368, 512)
self.bnf1 = nn.BatchNorm1d(64)
self.reluf1 = nn.ReLU(inplace=True)
# self.reluf1 = nn.Sigmoid()
# self.dropoutf1 = nn.Dropout(drop_prob)
self.fc2 = nn.Linear(64, 3)
# self.bnf2 = nn.BatchNorm1d(3)
# self.reluf2 = nn.ReLU(inplace=True)
# self.dropoutf2 = nn.Dropout(.25)
# self.fc2 = nn.Linear(32, num_classes)
# self.softmax = nn.Softmax(dim=1)
def forward(self, A, B, C):
# print(A.shape)
out = torch.cat([A,B,C], dim=2)
# print(out.shape)
# exit()
out = self.pool1(self.dropout1(self.relu1(self.bn1(self.conv1(out)))))
out = self.pool2(self.dropout2(self.relu2(self.bn2(self.conv2(out)))))
out = self.pool3(self.dropout3(self.relu3(self.bn3(self.conv3(out)))))
out = self.pool4(self.dropout4(self.relu4(self.bn4(self.conv4(out)))))
out = self.pool5(self.dropout5(self.relu5(self.bn5(self.conv5(out)))))
# out = self.maxpool6(self.dropout6(self.relu6(self.bn6(self.conv6(out)))))
# out = self.maxpool7(self.dropout7(self.relu7(self.bn7(self.conv7(out)))))
# out = self.dropout1(self.relu1(self.bn1(self.conv1(out))))
# out = self.dropout2(self.relu2(self.bn2(self.conv2(out))))
# out = self.dropout3(self.relu3(self.bn3(self.conv3(out))))
# out = self.dropout4(self.relu4(self.bn4(self.conv4(out))))
# out = self.avgpool1(self.relu1(self.bn1(self.conv1(out))))
# out = self.avgpool2(self.relu2(self.bn2(self.conv2(out))))
# out = self.avgpool3(self.relu3(self.bn3(self.conv3(out))))
# out = self.avgpool4(self.relu4(self.bn4(self.conv4(out))))
# out = self.avgpool5(self.relu5(self.bn5(self.conv5(out))))
# out = self.avgpool6(self.relu6(self.bn6(self.conv6(out))))
# out = self.avgpool7(self.relu7(self.bn7(self.conv7(out))))
# print(out.shape)
out = out.view(out.shape[0], -1)
# print(out.shape)
# exit()
out = self.reluf1(self.bnf1(self.fc1(out)))
# out = self.dropoutf2(self.reluf2(self.bnf2(self.fc2(out))))
out = self.fc2(out)
# out = self.softmax(out)
return out