Stride setting error in DeformConv2D?
Closed this issue · 1 comments
smilewsw commented
In the line
self.conv_kernel = nn.Conv2d(inc, outc, kernel_size=kernel_size, stride=kernel_size, bias=bias)
why do you set the stride the same as kernel_size? Is it an error?
ChunhuanLin commented
No, it's designed to suit the slide window for features extracted from the deformable offsets. You can review the code here to better understand my purpose.