Pith. sign in

ToSA: Token Merging with Spatial Awareness

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it
abstract

Token merging has emerged as an effective strategy to accelerate Vision Transformers (ViT) by reducing computational costs. However, existing methods primarily rely on the visual token's feature similarity for token merging, overlooking the potential of integrating spatial information, which can serve as a reliable criterion for token merging in the early layers of ViT, where the visual tokens only possess weak visual information. In this paper, we propose ToSA, a novel token merging method that combines both semantic and spatial awareness to guide the token merging process. ToSA leverages the depth image as input to generate pseudo spatial tokens, which serve as auxiliary spatial information for the visual token merging process. With the introduced spatial awareness, ToSA achieves a more informed merging strategy that better preserves critical scene structure. Experimental results demonstrate that ToSA outperforms previous token merging methods across multiple benchmarks on visual and embodied question answering while largely reducing the runtime of the ViT, making it an efficient solution for ViT acceleration. The code will be available at: https://github.com/hsiangwei0903/ToSA

citation-role summary

background 1

citation-polarity summary

fields

cs.CV 1

years

2025 1

verdicts

CONDITIONAL 1

roles

background 1

polarities

unclear 1

representative citing papers

Warehouse Spatial Question Answering with LLM Agent

cs.CV · 2025-07-14 · conditional · novelty 4.0

An LLM agent equipped with lightweight distance and inclusion perception models achieved 95.86% accuracy on the 2025 AI City Challenge warehouse spatial QA benchmark, ranking first.

citing papers explorer

Showing 1 of 1 citing paper.

  • Warehouse Spatial Question Answering with LLM Agent cs.CV · 2025-07-14 · conditional · none · ref 6 · internal anchor

    An LLM agent equipped with lightweight distance and inclusion perception models achieved 95.86% accuracy on the 2025 AI City Challenge warehouse spatial QA benchmark, ranking first.