search

yxgeee / MMT

[ICLR-2020] Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person Re-identification.
https://yxgeee.github.io/projects/mmt
MIT License
469 stars 73 forks source link

About sampler #38

Closed signofyouth closed 3 years ago

signofyouth commented 3 years ago

Thank you for your great work. I'm very interested in your paper.
When trying your code, I found RandomMultipleGallerySampler sometimes gives an ID which contains only 1 instance in a batch. This is because 1 instance pushed in 'ret' at line 74 was remained even when continued at https://github.com/yxgeee/MMT/blob/master/mmt/utils/data/sampler.py#L93.
Would you check and fix it? Thank you.

yxgeee commented 3 years ago

Yes, but it is not a bug here. Does it cause any running issues in your experiments?

signofyouth commented 3 years ago

Thank you for your quick reply.
No, errors does not occur when running on your code. And the result does not change as far as I tried. However, it's a little uncomfortable that the sampling policy(4 instances per ID) would not always work.

Anyway, you're right it is not a bug here. Please close this issue.