空洞卷积的等效卷积核(感受野)⼤⼩:
ke = k + (k − 1)(r − 1)
k为原始卷积核⼤⼩,r为dia rate参数。
输出计算同卷积运算。
输出特征⼤⼩计算
out = ( in − F + 2 ∗ pading ) / stride +1
注意:这⾥的F就是上⾯的感受野,这个公式也适⽤于常规的卷积
例如在deeplabv3+中特征⾦字塔模块:
#dilations = [1, 6, 12, 18]#
self.aspp1 = _ASPPModule(inplanes, 256,1, padding=0, dilation=dilations[0], BatchNorm=BatchNorm)
self.aspp2 = _ASPPModule(inplanes, 256,3, padding=dilations[1], dilation=dilations[1], BatchNorm=BatchNorm)
self.aspp3 = _ASPPModule(inplanes, 256,3, padding=dilations[2], dilation=dilations[2],BatchNorm=BatchNorm)
self.aspp4 = _ASPPModule(inplanes, 256,3, padding=dilations[3],dilation=dilations[3], BatchNorm=BatchNorm)
以上四步输出尺⼨⼤⼩相同 256*256
参考:
https://www.jianshu.com/p/a74117aea9ad
https://blog.csdn.net/m0_45447650/article/details/125540126