Content-aware Token Sharing for Efficient Semantic Segmentation with Vision Transformers

TL;DR

This paper proposes Content-aware Token Sharing (CTS), a novel method that reduces tokens in Vision Transformer-based semantic segmentation, achieving up to 44% efficiency gain without sacrificing accuracy.

Contribution

It introduces a class-agnostic policy network for token sharing in semantic segmentation, addressing limitations of previous token reduction methods designed for classification.

Findings

Reduces tokens processed by up to 44%

Maintains segmentation quality across datasets

Effective with various ViT backbones and decoders

Abstract

This paper introduces Content-aware Token Sharing (CTS), a token reduction approach that improves the computational efficiency of semantic segmentation networks that use Vision Transformers (ViTs). Existing works have proposed token reduction approaches to improve the efficiency of ViT-based image classification networks, but these methods are not directly applicable to semantic segmentation, which we address in this work. We observe that, for semantic segmentation, multiple image patches can share a token if they contain the same semantic class, as they contain redundant information. Our approach leverages this by employing an efficient, class-agnostic policy network that predicts if image patches contain the same semantic class, and lets them share a token if they do. With experiments, we explore the critical design choices of CTS and show its effectiveness on the ADE20K, Pascal Context and Cityscapes datasets, various ViT backbones, and different segmentation decoders. With Content-aware Token Sharing, we are able to reduce the number of processed tokens by up to 44%, without diminishing the segmentation quality.