search

yfeng95 / PRNet

Joint 3D Face Reconstruction and Dense Alignment with Position Map Regression Network (ECCV 2018)
http://openaccess.thecvf.com/content_ECCV_2018/papers/Yao_Feng_Joint_3D_Face_ECCV_2018_paper.pdf
MIT License
4.96k stars 944 forks source link

per image predict time #131

Open bluesliuf opened 5 years ago

bluesliuf commented 5 years ago

Thank you for your sharing code. I want to know the test time of a single picture on the CPU and GPU. So I used your test script run_basics.py and modified 2 parts

  1. Only put a picture in the test folder (TestImages_AFLW2000)
  2. Annotated lines 119 and 121 in api.py https://github.com/YadiraF/PRNet/blob/master/api.py#L119. Finally, the 0.23s on the CPU is similar to yours, but the 70ms on the GPU is very different from your result of 9.8ms. Could you tell me what the reason may cause this?