Alibaba-MIIL / ASL

Official Pytorch Implementation of: "Asymmetric Loss For Multi-Label Classification"(ICCV, 2021) paper
MIT License
732 stars 102 forks source link

some confusions about computing p_m #28

Closed Holmes-GU closed 3 years ago

Holmes-GU commented 3 years ago

Hi: I am reading your paper and your source codes, there are some confusions:

  1. In the paper, p_m = max(p-m, 0), however, in your implementation, p_m = p + 0.05. Are there some insights? Thanks~
mrT23 commented 3 years ago

see https://github.com/Alibaba-MIIL/ASL/issues/10

Holmes-GU commented 3 years ago

Ok,thank you~