Pytorch实现姿态识别（二）——视频分析C3D的网络架构
C3D网络架构与结构图：
3D卷积与2D卷积的区别：其中多了一个时间维度
三维卷积与三维池化的理解：
① nn.Conv3d(3, 64, kernel_size=(3, 3, 3), padding=(1, 1, 1))
kernel_size=(3, 3, 3)：
第一个3：一共16帧，当前处理3帧；
第二、三个3：H,W的大小；
padding=(1, 1, 1)：三维、高、宽都填充1
② nn.MaxPool3d(kernel_size=(1, 2, 2), stride=(1, 2, 2))
kernel_size=(1, 2, 2), stride=(1, 2, 2)：
第一个1：时间序列上的长度为1，16帧数据经过pooling还是16；作者不希望刚开始的时候压缩在时间序列上的特征。
第二、三个2：H,W的长度为2，高、宽经pooling后变为原来的一半。
import torch
import torch.nn as nn
from mypath import Path
class C3D(nn.Module):
def __init__(self, num_classes, pretrained=False):
super(C3D, self).__init__()
self.conv1 = nn.Conv3d(3, 64, kernel_size=(3, 3, 3), padding=(1, 1, 1))#16x56x56x64
self.pool1 = nn.MaxPool3d(kernel_size=(1, 2, 2), stride=(1, 2, 2))#16x56x56x64
self.conv2 = nn.Conv3d(64, 128, kernel_size=(3, 3, 3), padding=(1, 1, 1))#16x56x56x128
self.pool2 = nn.MaxPool3d(kernel_size=(2, 2, 2), stride=(2, 2, 2))#8x28x28x128
self.conv3a = nn.Conv3d(128, 256, kernel_size=(3, 3, 3), padding=(1, 1, 1))#8x28x28x256
self.conv3b = nn.Conv3d(256, 256, kernel_size=(3, 3, 3), padding=(1, 1, 1))#8x28x28x256
self.pool3 = nn.MaxPool3d(kernel_size=(2, 2, 2), stride=(2, 2, 2))#4x14x14x256
self.conv4a = nn.Conv3d(256, 512, kernel_size=(3, 3, 3), padding=(1, 1, 1))#4x14x14x512
self.conv4b = nn.Conv3d(512, 512, kernel_size=(3, 3, 3), padding=(1, 1, 1))#4x14x14x512
self.pool4 = nn.MaxPool3d(kernel_size=(2, 2, 2), stride=(2, 2, 2))# 2x7x7x512
self.conv5a = nn.Conv3d(512, 512, kernel_size=(3, 3, 3), padding=(1, 1, 1))# 2x7x7x512
self.conv5b = nn.Conv3d(512, 512, kernel_size=(3, 3, 3), padding=(1, 1, 1))# 2x7x7x512
self.pool5 = nn.MaxPool3d(kernel_size=(2, 2, 2), stride=(2, 2, 2), padding=(0, 1, 1))#1x4x4x512
self.fc6 = nn.Linear(8192, 4096)
self.fc7 = nn.Linear(4096, 4096)
self.fc8 = nn.Linear(4096, num_classes)
self.dropout = nn.Dropout(p=0.5)
self.relu = nn.ReLU()
self.__init_weight()
if pretrained:
self.__load_pretrained_weights()
def forward(self, x):
# batch_size,channels,帧数,H,W
# x.size=[6,3,16,112,112]
x = self.relu(self.conv1(x))#x.size:[6,64,16,112,112]
x = self.pool1(x)
x = self.relu(self.conv2(x))
x = self.pool2(x)
x = self.relu(self.conv3a(x))
x = self.relu(self.conv3b(x))
x = self.pool3(x)
x = self.relu(self.conv4a(x))
x = self.relu(self.conv4b(x))
x = self.pool4(x)
x = self.relu(self.conv5a(x))
x = self.relu(self.conv5b(x))
x = self.pool5(x)# x.size:[1,512,1,4,4]
x = x.view(-1, 8192)
x = self.relu(self.fc6(x))# x.size:[6,8192]
x = self.dropout(x)
x = self.relu(self.fc7(x))
x = self.dropout(x)
logits = self.fc8(x)# [6,101]
return logits
def __load_pretrained_weights(self):
"""Initialiaze network."""
corresp_name = {
# Conv1
"features.0.weight": "conv1.weight",
"features.0.bias": "conv1.bias",
# Conv2
"features.3.weight": "conv2.weight",
"features.3.bias": "conv2.bias",
# Conv3a
"features.6.weight": "conv3a.weight",
"features.6.bias": "conv3a.bias",
# Conv3b
"features.8.weight": "conv3b.weight",
"features.8.bias": "conv3b.bias",
# Conv4a
"features.11.weight": "conv4a.weight",
"features.11.bias": "conv4a.bias",
# Conv4b
"features.13.weight": "conv4b.weight",
"features.13.bias": "conv4b.bias",
# Conv5a
"features.16.weight": "conv5a.weight",
"features.16.bias": "conv5a.bias",
# Conv5b
"features.18.weight": "conv5b.weight",
"features.18.bias": "conv5b.bias",
# fc6
"classifier.0.weight": "fc6.weight",
"classifier.0.bias": "fc6.bias",
# fc7
"classifier.3.weight": "fc7.weight",
"classifier.3.bias": "fc7.bias",
}
p_dict = torch.load(Path.model_dir())
s_dict = self.state_dict()
for name in p_dict:
if name not in corresp_name:
continue
s_dict[corresp_name[name]] = p_dict[name]
self.load_state_dict(s_dict)
def __init_weight(self):
for m in self.modules():
if isinstance(m, nn.Conv3d):
# n = m.kernel_size[0] * m.kernel_size[1] * m.out_channels
# m.weight.data.normal_(0, math.sqrt(2. / n))
torch.nn.init.kaiming_normal_(m.weight)
elif isinstance(m, nn.BatchNorm3d):
m.weight.data.fill_(1)
m.bias.data.zero_()
def get_1x_lr_params(model):
"""
This generator returns all the parameters for conv and two fc layers of the net.
"""
b = [model.conv1, model.conv2, model.conv3a, model.conv3b, model.conv4a, model.conv4b,
model.conv5a, model.conv5b, model.fc6, model.fc7]
for i in range(len(b)):
for k in b[i].parameters():
if k.requires_grad:
yield k
def get_10x_lr_params(model):
"""
This generator returns all the parameters for the last fc layer of the net.
"""
b = [model.fc8]
for j in range(len(b)):
for k in b[j].parameters():
if k.requires_grad:
yield k
if __name__ == "__main__":
inputs = torch.rand(1, 3, 16, 112, 112)
net = C3D(num_classes=101, pretrained=True)
outputs = net.forward(inputs)
print(outputs.size())