With AutoAugment(RandAugment) or CutMix ?

[ildoonet]

I have tried to train RegNet variants with strong augmentations, such as AutoAugment or CutMix.

But the performance can not be improved with them.

For example, I have reproduced the paper's result for RegNet-Y-400M but I got around top1 accuracy 69% with CutMix, which is way below the vanilla RegNet-Y-400M.

I also tried to train the RegNet-Y with longer epochs, but failed to improve the result.

Do you have any experience for the strong data augmentations?

[ir413]

Hi @ildoonet, thanks for sharing your findings. I don't have experience with strong data augmentations myself. However, we heard that several groups have managed to achieve strong results with RegNets using improved training settings (e.g. RegNetY-3.2GF with 82.0 top-1 accuracy reported here). So I believe that some combination of stronger settings should be able to give you better results. We also plan to release models with stronger settings in the future.