facebookresearch / facebookresearch/SlowFast

Training Kinetics with color augmentation

Open
#374 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
7.4k
Forks
1.3k
PR merge metrics
No merged PRs in 30d

Description

Hi Team,

I had a question about training AVA vs Kinetics.

During AVA training, you are also doing colour augmentation, but the same is not done during the Kinetics training process. Is there any specific reason for so that you encountered during experimentation? Shouldn't doing colour augmentation on Kinetics should yield greater accuracy theoretically?

Thanks.

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.