ACC-ViT : Atrous Convolution's Comeback in Vision Transformers

TL;DR

This paper introduces ACC-ViT, a hybrid vision transformer that combines regional and sparse attention through Atrous Attention, inspired by atrous convolution, achieving high accuracy with fewer parameters and versatility across tasks.

Contribution

The work proposes Atrous Attention to fuse local and global information in vision transformers and redesigns convolution blocks with atrous convolution, creating a versatile hybrid backbone.

Findings

Achieves ~84% accuracy on ImageNet-1K with fewer parameters.

Outperforms state-of-the-art MaxViT by 0.42% accuracy.

Effective across various tasks like medical imaging, detection, and contrastive learning.

Abstract

Transformers have elevated to the state-of-the-art vision architectures through innovations in attention mechanism inspired from visual perception. At present two classes of attentions prevail in vision transformers, regional and sparse attention. The former bounds the pixel interactions within a region; the latter spreads them across sparse grids. The opposing natures of them have resulted in a dilemma between either preserving hierarchical relation or attaining a global context. In this work, taking inspiration from atrous convolution, we introduce Atrous Attention, a fusion of regional and sparse attention, which can adaptively consolidate both local and global information, while maintaining hierarchical relations. As a further tribute to atrous convolution, we redesign the ubiquitous inverted residual convolution blocks with atrous convolution. Finally, we propose a generalized, hybrid vision transformer backbone, named ACC-ViT, following conventional practices for standard vision tasks. Our tiny version model achieves $\sim 84 \%$ accuracy on ImageNet-1K, with less than $28.5$ million parameters, which is $0.42\%$ improvement over state-of-the-art MaxViT while having $8.4\%$ less parameters. In addition, we have investigated the efficacy of ACC-ViT backbone under different evaluation settings, such as finetuning, linear probing, and zero-shot learning on tasks involving medical image analysis, object detection, and language-image contrastive learning. ACC-ViT is therefore a strong vision backbone, which is also competitive in mobile-scale versions, ideal for niche applications with small datasets.