The Right to Talk: An Audio-Visual Transformer Approach

TL;DR

This paper presents a novel Audio-Visual Transformer method for localizing and highlighting the main speaker in multi-speaker videos, effectively exploiting cross-modal and temporal relationships to improve speaker detection.

Contribution

It introduces a new Transformer-based approach that models audio-visual correlations and temporal dynamics for main speaker localization, along with a newly collected dataset.

Findings

Effective localization of main speakers in multi-speaker videos.

Improved accuracy over previous methods in audio-visual speaker detection.

First study to automatically localize main speakers in both audio and visual channels.

Abstract

Turn-taking has played an essential role in structuring the regulation of a conversation. The task of identifying the main speaker (who is properly taking his/her turn of speaking) and the interrupters (who are interrupting or reacting to the main speaker's utterances) remains a challenging task. Although some prior methods have partially addressed this task, there still remain some limitations. Firstly, a direct association of Audio and Visual features may limit the correlations to be extracted due to different modalities. Secondly, the relationship across temporal segments helping to maintain the consistency of localization, separation, and conversation contexts is not effectively exploited. Finally, the interactions between speakers that usually contain the tracking and anticipatory decisions about the transition to a new speaker are usually ignored. Therefore, this work introduces a new Audio-Visual Transformer approach to the problem of localization and highlighting the main speaker in both audio and visual channels of a multi-speaker conversation video in the wild. The proposed method exploits different types of correlations presented in both visual and audio signals. The temporal audio-visual relationships across spatial-temporal space are anticipated and optimized via the self-attention mechanism in a Transformerstructure. Moreover, a newly collected dataset is introduced for the main speaker detection. To the best of our knowledge, it is one of the first studies that is able to automatically localize and highlight the main speaker in both visual and audio channels in multi-speaker conversation videos.