本篇文章分享论文『Uniform Masking: Enabling MAE Pre-training for Pyramid-based Vision Transformers with Locality』，南理工&上海AI Lab提出Uniform Masking，为基于金字塔结构的视觉Transformer进行MAE预训练！