Active Speaker Detection as a Multi-Objective Optimization with Uncertainty-Based Multimodal Fusion - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2021

Active Speaker Detection as a Multi-Objective Optimization with Uncertainty-Based Multimodal Fusion

Abstract

It is now well established from a variety of studies that there is a significant benefit from combining video and audio data in detecting active speakers. However, either of the modalities can potentially mislead audiovisual fusion by inducing unreliable or deceptive information. This paper outlines active speaker detection as a multi-objective learning problem to leverage best of each modalities using a novel self-attention, uncertainty-based multimodal fusion scheme. Results obtained show that the proposed multi-objective learning architecture outperforms traditional approaches in improving both mAP and AUC scores. We further demonstrate that our fusion strategy surpasses, in active speaker detection, other modality fusion methods reported in various disciplines. We finally show that the proposed method significantly improves the state-of-the-art on the AVA-ActiveSpeaker dataset.

Dates and versions

hal-03345281 , version 1 (15-09-2021)

Identifiers

Cite

Baptiste Pouthier, Laurent Pilati, Leela Gudupudi, Charles Bouveyron, Frédéric Precioso. Active Speaker Detection as a Multi-Objective Optimization with Uncertainty-Based Multimodal Fusion. Interspeech 2021, Aug 2021, Brno, Czech Republic. pp.2381-2385, ⟨10.21437/Interspeech.2021-80⟩. ⟨hal-03345281⟩
78 View
0 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More