-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathmodel.py
203 lines (166 loc) · 8.19 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
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
import torch
import torch.nn as nn
from params import par
from torch.autograd import Variable
from torch.nn.init import kaiming_normal_, orthogonal_
import numpy as np
from torch.distributions.utils import broadcast_all, probs_to_logits, logits_to_probs, lazy_property, clamp_probs
import torch.nn.functional as F
def conv(batchNorm, in_planes, out_planes, kernel_size=3, stride=1, dropout=0):
if batchNorm:
return nn.Sequential(
nn.Conv2d(in_planes, out_planes, kernel_size=kernel_size, stride=stride, padding=(kernel_size-1)//2, bias=False),
nn.BatchNorm2d(out_planes),
nn.LeakyReLU(0.1, inplace=True),
nn.Dropout(dropout)#, inplace=True)
)
else:
return nn.Sequential(
nn.Conv2d(in_planes, out_planes, kernel_size=kernel_size, stride=stride, padding=(kernel_size-1)//2, bias=True),
nn.LeakyReLU(0.1, inplace=True),
nn.Dropout(dropout)#, inplace=True)
)
class DeepVO(nn.Module):
def __init__(self, imsize1, imsize2, batchNorm=True):
super(DeepVO,self).__init__()
# CNN
self.batchNorm = batchNorm
self.conv1 = conv(self.batchNorm, 6, 64, kernel_size=7, stride=2, dropout=par.conv_dropout[0])
self.conv2 = conv(self.batchNorm, 64, 128, kernel_size=5, stride=2, dropout=par.conv_dropout[1])
self.conv3 = conv(self.batchNorm, 128, 256, kernel_size=5, stride=2, dropout=par.conv_dropout[2])
self.conv3_1 = conv(self.batchNorm, 256, 256, kernel_size=3, stride=1, dropout=par.conv_dropout[3])
self.conv4 = conv(self.batchNorm, 256, 512, kernel_size=3, stride=2, dropout=par.conv_dropout[4])
self.conv4_1 = conv(self.batchNorm, 512, 512, kernel_size=3, stride=1, dropout=par.conv_dropout[5])
self.conv5 = conv(self.batchNorm, 512, 512, kernel_size=3, stride=2, dropout=par.conv_dropout[6])
self.conv5_1 = conv(self.batchNorm, 512, 512, kernel_size=3, stride=1, dropout=par.conv_dropout[7])
self.conv6 = conv(self.batchNorm, 512, 1024, kernel_size=3, stride=2, dropout=par.conv_dropout[8])
# Comput the shape based on diff image size
__tmp = Variable(torch.zeros(1, 6, imsize1, imsize2))
__tmp = self.encode_image(__tmp)
# RNN
self.rnn = nn.LSTM(
input_size=int(np.prod(__tmp.size())),
hidden_size=par.rnn_hidden_size,
num_layers=2,
dropout=par.rnn_dropout_between,
batch_first=True)
self.rnn_drop_out = nn.Dropout(par.rnn_dropout_out)
self.linear = nn.Linear(in_features=par.rnn_hidden_size, out_features=6)
# Initilization
for m in self.modules():
if isinstance(m, nn.Conv2d) or isinstance(m, nn.ConvTranspose2d) or isinstance(m, nn.Linear):
kaiming_normal_(m.weight.data)
if m.bias is not None:
m.bias.data.zero_()
elif isinstance(m, nn.LSTM):
# layer 1
kaiming_normal_(m.weight_ih_l0) #orthogonal_(m.weight_ih_l0)
kaiming_normal_(m.weight_hh_l0)
m.bias_ih_l0.data.zero_()
m.bias_hh_l0.data.zero_()
# Set forget gate bias to 1 (remember)
n = m.bias_hh_l0.size(0)
start, end = n//4, n//2
m.bias_hh_l0.data[start:end].fill_(1.)
# layer 2
kaiming_normal_(m.weight_ih_l1) #orthogonal_(m.weight_ih_l1)
kaiming_normal_(m.weight_hh_l1)
m.bias_ih_l1.data.zero_()
m.bias_hh_l1.data.zero_()
n = m.bias_hh_l1.size(0)
start, end = n//4, n//2
m.bias_hh_l1.data[start:end].fill_(1.)
elif isinstance(m, nn.BatchNorm2d):
m.weight.data.fill_(1)
m.bias.data.zero_()
def load_Flownet(self):
# Load the pre-trained FlowNet
pretrained_w = torch.load(par.pretrained_flownet, map_location='cpu')
# Load FlowNet weights pretrained with FlyingChairs
# NOTE: the pretrained model assumes image rgb values in range [-0.5, 0.5]
if par.pretrained_flownet and not par.resume:
# Use only conv-layer-part of FlowNet as CNN for DeepVO
model_dict = self.state_dict()
update_dict = {k: v for k, v in pretrained_w['state_dict'].items() if k in model_dict}
model_dict.update(update_dict)
self.load_state_dict(model_dict)
def forward(self, x, prev=None):
# x: (batch, seq_len, channel, width, height)
# stack_image
x = torch.cat((x[:, :-1], x[:, 1:]), dim=2)
batch_size = x.size(0)
seq_len = x.size(1)
# CNN
x = x.view(batch_size*seq_len, x.size(2), x.size(3), x.size(4))
x = self.encode_image(x)
x = x.view(batch_size, seq_len, -1)
# RNN
out, hc = self.rnn(x) if not prev else self.rnn(x, prev)
out = self.rnn_drop_out(out)
pose = self.linear(out)
angle = pose[:, :, :3]
trans = pose[:, :, 3:]
return angle, trans, hc
def encode_image(self, x):
out_conv2 = self.conv2(self.conv1(x))
out_conv3 = self.conv3_1(self.conv3(out_conv2))
out_conv4 = self.conv4_1(self.conv4(out_conv3))
out_conv5 = self.conv5_1(self.conv5(out_conv4))
out_conv6 = self.conv6(out_conv5)
return out_conv6
def weight_parameters(self):
return [param for name, param in self.named_parameters() if 'weight' in name]
def bias_parameters(self):
return [param for name, param in self.named_parameters() if 'bias' in name]
def get_loss(self, x, y, prev=None):
angle, trans, _ = self.forward(x, prev=prev)
angle_loss = torch.nn.functional.mse_loss(angle, y[:,:,:3])
translation_loss = torch.nn.functional.mse_loss(trans, y[:,:,3:])
loss = 100 * angle_loss + translation_loss
return loss, angle_loss, translation_loss
def step(self, x, y, optimizer, prev=None):
optimizer.zero_grad()
loss, angle_loss, translation_loss = self.get_loss(x, y, prev=prev)
loss.backward()
optimizer.step()
return loss, angle_loss, translation_loss
def train_net(self, dataloader, optimizer):
self.train()
loss_ang_mean_train = 0
loss_trans_mean_train = 0
iter_num = 0
for t_x, t_y in dataloader:
t_x = t_x.to(par.device)
t_y = t_y.to(par.device)
loss, angle_loss, translation_loss = self.step(t_x, t_y, optimizer)
loss = loss.data.cpu().numpy()
angle_loss = angle_loss.data.cpu().numpy()
translation_loss = translation_loss.data.cpu().numpy()
loss_ang_mean_train += float(angle_loss) * 100
loss_trans_mean_train += float(translation_loss)
iter_num += 1
if iter_num % 20 == 0:
message = f'Iteration: {iter_num}, Loss: {loss:.3f}, angle: {100*angle_loss:.4f}, trans: {translation_loss:.3f}'
f = open(par.record_path, 'a')
f.write(message+'\n')
print(message)
loss_ang_mean_train /= len(dataloader)
loss_trans_mean_train /= len(dataloader)
loss_mean_train = loss_ang_mean_train + loss_trans_mean_train
return loss_mean_train, loss_ang_mean_train, loss_trans_mean_train
def valid_net(self, dataloader):
self.eval()
loss_ang_mean_valid = 0
loss_trans_mean_valid = 0
loss_yaw_mean_valid = 0
for v_x, v_y in dataloader:
v_x = v_x.to(par.device)
v_y = v_y.to(par.device)
loss, angle_loss, translation_loss = self.get_loss(v_x, v_y)
loss = loss.data.cpu().numpy()
angle_loss = angle_loss.data.cpu().numpy()
translation_loss = translation_loss.data.cpu().numpy()
loss_ang_mean_valid += float(angle_loss) * 100
loss_trans_mean_valid += float(translation_loss)
loss_mean_valid = loss_ang_mean_valid + loss_trans_mean_valid
return loss_mean_valid, loss_ang_mean_valid, loss_trans_mean_valid