STHG: Spatial-Temporal Heterogeneous Graph Learning for Advanced Audio-Visual Diarization

TL;DR

The paper presents STHG, a novel spatial-temporal heterogeneous graph learning method for audio-visual diarization that jointly models all speakers, including the camera wearer, achieving state-of-the-art results in the Ego4D Challenge 2023.

Contribution

Introduces a unified heterogeneous graph learning framework for audio-visual diarization that outperforms previous methods and includes all speakers without separate components.

Findings

Achieved 61.1% DER on Ego4D test set.

Outperformed all baselines and last year's winner.

Secured 1st place in the Ego4D Challenge 2023.

Abstract

This report introduces our novel method named STHG for the Audio-Visual Diarization task of the Ego4D Challenge 2023. Our key innovation is that we model all the speakers in a video using a single, unified heterogeneous graph learning framework. Unlike previous approaches that require a separate component solely for the camera wearer, STHG can jointly detect the speech activities of all people including the camera wearer. Our final method obtains 61.1% DER on the test set of Ego4D, which significantly outperforms all the baselines as well as last year's winner. Our submission achieved 1st place in the Ego4D Challenge 2023. We additionally demonstrate that applying the off-the-shelf speech recognition system to the diarized speech segments by STHG produces a competitive performance on the Speech Transcription task of this challenge.