Seeing the Context: Rich Visual Context-Aware Speech Recognition via Multimodal Reasoning

TL;DR

This paper introduces VASR, a multimodal reasoning framework for audio-visual speech recognition that leverages rich visual context and explicit reasoning to improve accuracy and address modality over-reliance.

Contribution

It proposes AV-CoT, a novel reasoning approach that explicitly grounds audio and visual signals, and provides a new dataset and pipeline for context-aware AVSR.

Findings

AV-CoT mitigates single-modality dominance effectively.

Achieves state-of-the-art results in context-aware AVSR.

Open-sourced data pipeline and test set for further research.

Abstract

Audio-visual speech recognition (AVSR) is an extension of ASR that incorporates visual signals. Current AVSR approaches primarily focus on lip motion, largely overlooking rich context present in the video such as speaking scene and on-screen text. To tackle such CAVSR (AVSR including rich visual Context), we propose VASR designed to "see" and reason the visual context to improve speech recognition. Specifically, we construct an Audio-Visual Chain-of-Thought (AV-CoT) that explicitly enforces intermediate cross-modal grounding between acoustic signals and visual evidence. This evidence-driven reasoning mitigates the "single-modality dominance" problem, where models either over-rely on visual context or fail to utilize it. Besides, to address the data scarcity, we construct and release a corresponding data pipeline and test set. Experiments show that AV-CoT effectively mitigates the single-modality dominance, achieving state-of-the-art performance in CAVSR. The project is open-sourced.