Question to authors: have you tested image augmentations in the training pipeline?

[FraPochetti]

Hi, first of all thanks a lot for the paper and the code. @PeikeLi I was wondering, have you tested using any image augmentations strategy in the training pipeline? Looking at your dataset, it doesn't seem like you are applying any. To be fair, LIP is large enough to probably make any augmentation redundant. Just curious though.