1 paper
Hexiang Bai, Hanyang Xu, Xiaoxue Li +4
Vision Transformers (ViTs) have shown immense potential in medical image analysis. However, standard pre-training via global image classification suffers from spatial collapse, whe…