Fetching the paper…

Efficient Self-supervised Vision Pretraining with Local Masked Reconstruction · Around