search

LiMeng95 / MultiPoseNet.pytorch

pytorch implementation of MultiPoseNet (ECCV 2018, Muhammed Kocabas et al.)
196 stars 32 forks source link

Maybe some bugs in network? #31

Closed H-Liu1997 closed 5 years ago

H-Liu1997 commented 5 years ago
  1. several layers don't using relu in forward function. here Are these layers no need to using relu in the FPN structure? in the multipose paper, they didn't say they using relu in some parts, but for smooth, they said using 3*3 conv and relu, it seems you forget it.
  2. in the paper, they using two 3*3 conv to do the transform form m2 to k2, in your model there are only one.

Maybe I have some misunderstanding portion, sorry to bother you and looking forward to your reply.