austingg / MobileNet-v2-caffe

MobileNet-v2 experimental network description for caffe

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

The model complexity(MAdds) is bigger than that in the paper

TaojiannanYang opened this issue · comments

Hi, I compute the the MAdds of your .prototxt, the MAdds is 313M, which is bigger than the MAdds stated in the paper(300M). I can not figure out where went wrong. I think my computation is right(I have used the script to compute other models), and I think your implementation is also right. Do you have any idea about this? Have you computed the MAdds of your implementation? Thanks in advance.

@TaojiannanYang sorry for late. I have checked the tf official mobilenet. The first bottleneck sequence whose expand ratio is 1, have only depthwise conv and conv linear, no conv 1x1 expand. And the missing conv 1x1's MAdd is about 13M ( 112x112x1x1x32x32). However, this detailed is not mentioned in the paper.