Masked Autoencoders Are Scalable Vision Learners

Masked Autoencoders Are Scalable Vision Learners.pdf