Lightweight Wasserstein Audio-Visual Model for Unified Speech Enhancement and Separation

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Park, Jisoo, Lee, Seonghak, Kim, Guisik, Kim, Taewoo, Kwon, Junseok
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866915659439407104
author Park, Jisoo
Lee, Seonghak
Kim, Guisik
Kim, Taewoo
Kwon, Junseok
author_facet Park, Jisoo
Lee, Seonghak
Kim, Guisik
Kim, Taewoo
Kwon, Junseok
contents Speech Enhancement (SE) and Speech Separation (SS) have traditionally been treated as distinct tasks in speech processing. However, real-world audio often involves both background noise and overlapping speakers, motivating the need for a unified solution. While recent approaches have attempted to integrate SE and SS within multi-stage architectures, these approaches typically involve complex, parameter-heavy models and rely on supervised training, limiting scalability and generalization. In this work, we propose UniVoiceLite, a lightweight and unsupervised audio-visual framework that unifies SE and SS within a single model. UniVoiceLite leverages lip motion and facial identity cues to guide speech extraction and employs Wasserstein distance regularization to stabilize the latent space without requiring paired noisy-clean data. Experimental results demonstrate that UniVoiceLite achieves strong performance in both noisy and multi-speaker scenarios, combining efficiency with robust generalization. The source code is available at https://github.com/jisoo-o/UniVoiceLite.
format Preprint
id arxiv_https___arxiv_org_abs_2512_06689
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Lightweight Wasserstein Audio-Visual Model for Unified Speech Enhancement and Separation
Park, Jisoo
Lee, Seonghak
Kim, Guisik
Kim, Taewoo
Kwon, Junseok
Computer Vision and Pattern Recognition
Audio and Speech Processing
Speech Enhancement (SE) and Speech Separation (SS) have traditionally been treated as distinct tasks in speech processing. However, real-world audio often involves both background noise and overlapping speakers, motivating the need for a unified solution. While recent approaches have attempted to integrate SE and SS within multi-stage architectures, these approaches typically involve complex, parameter-heavy models and rely on supervised training, limiting scalability and generalization. In this work, we propose UniVoiceLite, a lightweight and unsupervised audio-visual framework that unifies SE and SS within a single model. UniVoiceLite leverages lip motion and facial identity cues to guide speech extraction and employs Wasserstein distance regularization to stabilize the latent space without requiring paired noisy-clean data. Experimental results demonstrate that UniVoiceLite achieves strong performance in both noisy and multi-speaker scenarios, combining efficiency with robust generalization. The source code is available at https://github.com/jisoo-o/UniVoiceLite.
title Lightweight Wasserstein Audio-Visual Model for Unified Speech Enhancement and Separation
topic Computer Vision and Pattern Recognition
Audio and Speech Processing
url https://arxiv.org/abs/2512.06689