-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathflownet1s.py
175 lines (136 loc) · 6.91 KB
/
flownet1s.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
# Author: Jochen Gast <jochen.gast@visinf.tu-darmstadt.de>
import torch
import torch.nn as nn
import torch.nn.functional as tf
from contrib import weight_init
from contrib.interpolation import resize2d_as
from models import factory
def conv(in_planes, out_planes, kernel_size, stride, pad, nonlinear, bias):
if nonlinear:
return nn.Sequential(
nn.Conv2d(
in_planes, out_planes, kernel_size=kernel_size,
stride=stride, padding=pad, bias=bias),
nn.LeakyReLU(0.1, inplace=True)
)
else:
return nn.Conv2d(
in_planes, out_planes, kernel_size=kernel_size, stride=stride, padding=pad, bias=bias)
def deconv(in_planes, out_planes, kernel_size, stride, pad, nonlinear, bias):
if nonlinear:
return nn.Sequential(
nn.ConvTranspose2d(
in_planes, out_planes, kernel_size=kernel_size,
stride=stride, padding=pad, bias=bias),
nn.LeakyReLU(0.1, inplace=True)
)
else:
return nn.ConvTranspose2d(
in_planes, out_planes, kernel_size=kernel_size, stride=stride, padding=pad, bias=bias)
def concatenate_as(tensor_list, tensor_as, dim, mode="bilinear"):
tensor_list = [resize2d_as(x, tensor_as, mode=mode) for x in tensor_list]
return torch.cat(tensor_list, dim=dim)
def resize_tensors_as(tensor_list, tensor_as, mode="bilinear", align_corners=True):
if isinstance(tensor_list, torch.Tensor):
return resize2d_as(tensor_list, tensor_as, mode=mode, align_corners=align_corners)
else:
return [resize2d_as(x, tensor_as, mode=mode, align_corners=align_corners) for x in tensor_list]
def upsample2d_as(inputs, target_as, mode="bilinear", align_corners=True):
_, _, h, w = target_as.size()
return tf.interpolate(inputs, [h, w], mode=mode, align_corners=align_corners)
def upsample_flow_as(inputs, target_as, mode="bilinear", align_corners=True):
_, _, h, w = target_as.size()
factor = h / inputs.size(2)
return factor * tf.upsample(inputs, [h, w], mode=mode, align_corners=align_corners)
class FlowNetS(nn.Module):
def __init__(self, args, num_pred=2):
super().__init__()
self.args = args
self.num_pred = num_pred
def make_conv(in_planes, out_planes, kernel_size, stride):
pad = kernel_size // 2
return conv(in_planes, out_planes, kernel_size=kernel_size,
stride=stride, pad=pad, nonlinear=True, bias=True)
self.conv1 = make_conv(6, 64, kernel_size=7, stride=2)
self.conv2 = make_conv(64, 128, kernel_size=5, stride=2)
self.conv3 = make_conv(128, 256, kernel_size=5, stride=2)
self.conv3_1 = make_conv(256, 256, kernel_size=3, stride=1)
self.conv4 = make_conv(256, 512, kernel_size=3, stride=2)
self.conv4_1 = make_conv(512, 512, kernel_size=3, stride=1)
self.conv5 = make_conv(512, 512, kernel_size=3, stride=2)
self.conv5_1 = make_conv(512, 512, kernel_size=3, stride=1)
self.conv6 = make_conv(512, 1024, kernel_size=3, stride=2)
self.conv6_1 = make_conv(1024, 1024, kernel_size=3, stride=1)
def make_deconv(in_planes, out_planes):
return deconv(in_planes, out_planes, kernel_size=4, stride=2, pad=1,
nonlinear=True, bias=False)
self.deconv5 = make_deconv(1024, 512)
self.deconv4 = make_deconv(1024 + num_pred, 256)
self.deconv3 = make_deconv(768 + num_pred, 128)
self.deconv2 = make_deconv(384 + num_pred, 64)
def make_predict(in_planes, out_planes):
return conv(in_planes, out_planes, kernel_size=3, stride=1, pad=1,
nonlinear=False, bias=True)
self.predict_flow6 = make_predict(1024, num_pred)
self.predict_flow5 = make_predict(1024 + num_pred, num_pred)
self.predict_flow4 = make_predict(768 + num_pred, num_pred)
self.predict_flow3 = make_predict(384 + num_pred, num_pred)
self.predict_flow2 = make_predict(192 + num_pred, num_pred)
def make_upsample(in_planes, out_planes):
return deconv(in_planes, out_planes, kernel_size=4, stride=2, pad=1,
nonlinear=False, bias=False)
self.upsample_flow6_to_5 = make_upsample(num_pred, num_pred)
self.upsample_flow5_to_4 = make_upsample(num_pred, num_pred)
self.upsample_flow4_to_3 = make_upsample(num_pred, num_pred)
self.upsample_flow3_to_2 = make_upsample(num_pred, num_pred)
weight_init.msra_(self.modules(), mode='fan_in', nonlinearity='leaky_relu')
def forward(self, inputs):
conv1 = self.conv1(inputs)
conv2 = self.conv2(conv1)
conv3_1 = self.conv3_1(self.conv3(conv2))
conv4_1 = self.conv4_1(self.conv4(conv3_1))
conv5_1 = self.conv5_1(self.conv5(conv4_1))
conv6_1 = self.conv6_1(self.conv6(conv5_1))
predict_flow6 = self.predict_flow6(conv6_1)
upsampled_flow6_to_5 = self.upsample_flow6_to_5(predict_flow6)
deconv5 = self.deconv5(conv6_1)
concat5 = concatenate_as((conv5_1, deconv5, upsampled_flow6_to_5), conv5_1, dim=1)
predict_flow5 = self.predict_flow5(concat5)
upsampled_flow5_to_4 = self.upsample_flow5_to_4(predict_flow5)
deconv4 = self.deconv4(concat5)
concat4 = concatenate_as((conv4_1, deconv4, upsampled_flow5_to_4), conv4_1, dim=1)
predict_flow4 = self.predict_flow4(concat4)
upsampled_flow4_to_3 = self.upsample_flow4_to_3(predict_flow4)
deconv3 = self.deconv3(concat4)
concat3 = concatenate_as((conv3_1, deconv3, upsampled_flow4_to_3), conv3_1, dim=1)
predict_flow3 = self.predict_flow3(concat3)
upsampled_flow3_to_2 = self.upsample_flow3_to_2(predict_flow3)
deconv2 = self.deconv2(concat3)
concat2 = concatenate_as((conv2, deconv2, upsampled_flow3_to_2), conv2, dim=1)
predict_flow2 = self.predict_flow2(concat2)
if self.training:
return predict_flow2, predict_flow3, predict_flow4, predict_flow5, predict_flow6
else:
return predict_flow2
class FlowNet1S(nn.Module):
def __init__(self, args, div_flow=1.0):
super().__init__()
self.flownets = FlowNetS(args)
self.div_flow = div_flow
def forward(self, input_dict):
im1 = input_dict['input1']
im2 = input_dict['input2']
inputs = torch.cat((im1, im2), dim=1)
output_dict = {}
if self.training:
flow2, flow3, flow4, flow5, flow6 = self.flownets(inputs)
output_dict['flow2'] = flow2
output_dict['flow3'] = flow3
output_dict['flow4'] = flow4
output_dict['flow5'] = flow5
output_dict['flow6'] = flow6
else:
flow2 = self.flownets(inputs)
output_dict['flow1'] = upsample2d_as(flow2, im1, mode="bilinear") / self.div_flow
return output_dict
factory.register("FlowNet1S", FlowNet1S)