Long-Tailed-Recognition.pytorch icon indicating copy to clipboard operation
Long-Tailed-Recognition.pytorch copied to clipboard

Cifar Validation set

Open sxontheway opened this issue 3 years ago • 1 comments

Hello. I found in ImbalanceCIFAR.py, the validation set and the test set seem to be the same. Is it better to use a balanced training set (50000 frames) as the validation set instead of using the 10000 frame test set for validation? Thanks!

sxontheway avatar Apr 25 '21 11:04 sxontheway

Thank you for your advice. We noticed the same problem when we were working on Imbalanced CIFAR. However, since this split is adopted by all the previous methods, we have to use the same split for fair comparisons.

KaihuaTang avatar Jun 09 '21 06:06 KaihuaTang