论文《Fast Online Object Tracking and Segmentation- A Unifying Approach》项目代码解读2

1. Resnet.py

预定义的resnet网络代码

2. Rpn.py

DepthCorr最孪生网络最后一步过程实现，如下图黑框部分，类中定义三个网络，对应关系如图黑框黑体字标出。模板分支的特征提取结果作为卷积核对搜索分支的特征再次卷积。可参考SiamFC，SiamRPN。

class DepthCorr(nn.Module):    
	def __init__(self, in_channels, hidden, out_channels, kernel_size=3):        
		super(DepthCorr, self).__init__()        
		# adjust layer for asymmetrical features        
		self.conv_kernel = nn.Sequential(nn.Conv2d(in_channels, hidden, kernel_size=kernel_size, bias=False),nn.BatchNorm2d(hidden), nn.ReLU(inplace=True),)        
		self.conv_search = nn.Sequential(nn.Conv2d(in_channels, hidden, kernel_size=kernel_size, bias=False),nn.BatchNorm2d(hidden),nn.ReLU(inplace=True),)
		self.head = nn.Sequential(nn.Conv2d(hidden, hidden, kernel_size=1, bias=False),nn.BatchNorm2d(hidden),nn.ReLU(inplace=True),nn.Conv2d(hidden, out_channels, kernel_size=1))
	def forward_corr(self, kernel, input):       	
    		kernel = self.conv_kernel(kernel)        
    		input = self.conv_search(input)        
    		feature = conv2d_dw_group(input, kernel)        
    		return feature
	def forward(self, kernel, search):        
		feature = self.forward_corr(kernel, search)        
		out = self.head(feature)        
		return out

3.track_config.py

SiamMask网络超参数设置

。。。