Adaptive Attention Link-based Regularization for Vision Transformers

Heegon Jin, Jongwon Choi

Keywords: Vision transformers, Knowledge transfer, Knowledge distillation

TL;DR: The paper proposes a regularization method for vision transformers designed to make the attention maps contain similar structure to activation maps of CNNs.

Abstract: Although transformer networks are recently employed in the various vision tasks with the outperforming performance, large training data and a lengthy training time are required to train a model to disregard an inductive bias. Using trainable links between the channel-wise spatial attention of a pre-trained Convolutional Neural Network (CNN) and the attention head of Vision Transformers (ViT), we present a regularization technique to improve the training efficiency of Vision Transformers (ViT). The trainable links are referred to as the attention augmentation module, which is trained simultaneously with ViT, boosting the training of ViT and allowing it to avoid the overfitting issue caused by a lack of data. From the trained attention augmentation module, we can extract the relevant relationship between each CNN activation map and each ViT attention head, and based on this, we also propose an advanced attention augmentation module. Consequently, even with a small amount of data, the suggested method considerably improves the performance of ViT while achieving faster convergence during training.