ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases

Publication
arXiv preprint arXiv:2103.10697

Related