https://casualganpapers.com/self-supervised-large-scale-pretraining-vision-transformers/MAE-explained.html
Check out this paper summary