Jing Zhang

Thanks for the advice, feel free to post a pull requst.

The original repos @eyebies mentions doesn't work for openpose models. So I forked one and made the improvements -> https://github.com/jing-vision/pytorch-caffe-darknet-convert/blob/master/caffe2darknet.py I can confirm it works for openpose caffe models.

What's your GPU type?

能否给出更多信息？

感谢告知，我想想怎么 fix

+1

Check my project https://github.com/jing-vision/lightnet/tree/master/feature-viz

Yes, it should

I've answered in https://github.com/jing-vision/lightnet/issues/17#issuecomment-1201139827

COOL