/
training.py
135 lines (114 loc) · 4.2 KB
/
training.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
from __future__ import print_function
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
import torchvision
import matplotlib.pyplot as plt
import numpy as np
from stn import STNet, stn_train, stn_test
from stem import Stem
from visualise import imshow, save, plot_pores
from texture import Texture
from minutiae import Minutiae1a, Minutiae1b
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
class TextureNet(nn.Module):
def __init__(self):
super(TextureNet, self).__init__()
self.fc = nn.Linear(256, 1)
self.feature1 = nn.Sequential(
STNet(),
Stem(),
Texture()
)
self.inp = 0
def forward(self, x):
x11 = self.feature1(x)
x1 = x11.view(-1, 256)
x1 = self.fc(x1)
self.inp = x1
return x1, x11
class Minutiae2aNet(nn.Module):
def __init__(self):
super(Minutiae2aNet, self).__init__()
self.feature2a = nn.Sequential(
STNet(),
Stem(),
Minutiae1a()
)
def forward(self, x):
x2a = self.feature2a(x)
return x2a
class Minutiae2bNet(nn.Module):
def __init__(self):
super(Minutiae2bNet, self).__init__()
self.fc = nn.Linear(256, 1)
self.feature2b = nn.Sequential(
STNet(),
Stem(),
Minutiae1b()
)
self.inp = 0
def forward(self, x):
x2b2 = self.feature2b(x)
x2b = x2b2.view(-1, 256)
x2b = self.fc(x2b)
self.inp = x2b
return x2b, x2b2
model1 = TextureNet().to(device)
model2a = Minutiae2aNet().to(device)
model2b = Minutiae2bNet().to(device)
criterion1 = nn.CrossEntropyLoss()
criterion2a = nn.MSELoss()
criterion2b = nn.CrossEntropyLoss()
optimizer1 = optim.SGD(model1.parameters(), lr=0.001, momentum=0.9)
optimizer2a = optim.SGD(model2a.parameters(), lr=0.001, momentum=0.9)
optimizer2b = optim.SGD(model2b.parameters(), lr=0.001, momentum=0.9)
Transform = torchvision.transforms.Compose([torchvision.transforms.Grayscale(),
torchvision.transforms.Resize((132,132)),
torchvision.transforms.ToTensor()])
train_dataset = torchvision.datasets.ImageFolder(root='./imgs/',
transform=Transform)
#test_dataset = torchvision.datasets.ImageFolder(root='./imgs/test/',
# transform=Transform)
train_loader=torch.utils.data.DataLoader(train_dataset, batch_size=1, shuffle=True)
#test_loader=torch.utils.data.DataLoader(test_dataset, batch_size=1, shuffle=False)
#train_loader_org=torch.utils.data.DataLoader(train_dataset, batch_size=1, shuffle=False)
#
#stn_test(train_loader, 'shuffled')
#
#trainloader=torch.utils.data.DataLoader(torchvision.datasets.ImageFolder(root='./shuffled/',
# transform=Transform), batch_size=1, shuffle=True)
#
#trainloader_org=torch.utils.data.DataLoader(torchvision.datasets.ImageFolder(root='./aligned/original',
# transform=Transform), batch_size=1, shuffle=False)
# Train the model
#for j in range (10):
# model1.train()
# model2a.train()
# model2b.train()
#
# for i, data in enumerate(trainloader):
# inputs, target = data[0].to(device), data[1].to(device)
# target_f = torch.Tensor.float(target)
#
# optimizer1.zero_grad()
# optimizer2a.zero_grad()
# optimizer2b.zero_grad()
#
# output1 = model1(inputs)[0]
# output2a = model2a(inputs)
# output2b = model2b(inputs)[0]
#
# loss = (criterion1(F.softmax(output1), target) + criterion2a(output2a, target_f) +
# criterion2b(F.softmax(output2b), target))
# loss.backward()
#
# optimizer1.step()
# optimizer2a.step()
# optimizer2b.step()
#
# print(j, i, loss.item())
#torch.save(model1, 'trained1.pth')
#torch.save(model2a, 'trained2a.pth')
#torch.save(model2b, 'trained2b.pth')