search

wzmsltw / BSN-boundary-sensitive-network.pytorch

Codes of our paper: "BSN: Boundary Sensitive Network for Temporal Action Proposal Generation"
248 stars 58 forks source link

What's up with sigmoid loss multiplier? #18

Closed cinjon closed 5 years ago

cinjon commented 5 years ago

In this repo, it is .01. In the tensorflow rep, it is .1. Which (of either) did you actually use?

wzmsltw commented 5 years ago

This number is used to prevent nan. So both number can work.