Johns Hohns
LeViT: A Vision Transformer in ConvNet’s Clothing for Faster Inference (ICCV2021)