SEF-PNet: Speaker Encoder-Free Personalized Speech Enhancement with Local and Global Contexts Aggregation

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Huang, Ziling, Guan, Haixin, Wei, Haoran, Long, Yanhua
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866909461178744832
author Huang, Ziling
Guan, Haixin
Wei, Haoran
Long, Yanhua
author_facet Huang, Ziling
Guan, Haixin
Wei, Haoran
Long, Yanhua
contents Personalized speech enhancement (PSE) methods typically rely on pre-trained speaker verification models or self-designed speaker encoders to extract target speaker clues, guiding the PSE model in isolating the desired speech. However, these approaches suffer from significant model complexity and often underutilize enrollment speaker information, limiting the potential performance of the PSE model. To address these limitations, we propose a novel Speaker Encoder-Free PSE network, termed SEF-PNet, which fully exploits the information present in both the enrollment speech and noisy mixtures. SEF-PNet incorporates two key innovations: Interactive Speaker Adaptation (ISA) and Local-Global Context Aggregation (LCA). ISA dynamically modulates the interactions between enrollment and noisy signals to enhance the speaker adaptation, while LCA employs advanced channel attention within the PSE encoder to effectively integrate local and global contextual information, thus improving feature learning. Experiments on the Libri2Mix dataset demonstrate that SEF-PNet significantly outperforms baseline models, achieving state-of-the-art PSE performance.
format Preprint
id arxiv_https___arxiv_org_abs_2501_11274
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SEF-PNet: Speaker Encoder-Free Personalized Speech Enhancement with Local and Global Contexts Aggregation
Huang, Ziling
Guan, Haixin
Wei, Haoran
Long, Yanhua
Audio and Speech Processing
Sound
Personalized speech enhancement (PSE) methods typically rely on pre-trained speaker verification models or self-designed speaker encoders to extract target speaker clues, guiding the PSE model in isolating the desired speech. However, these approaches suffer from significant model complexity and often underutilize enrollment speaker information, limiting the potential performance of the PSE model. To address these limitations, we propose a novel Speaker Encoder-Free PSE network, termed SEF-PNet, which fully exploits the information present in both the enrollment speech and noisy mixtures. SEF-PNet incorporates two key innovations: Interactive Speaker Adaptation (ISA) and Local-Global Context Aggregation (LCA). ISA dynamically modulates the interactions between enrollment and noisy signals to enhance the speaker adaptation, while LCA employs advanced channel attention within the PSE encoder to effectively integrate local and global contextual information, thus improving feature learning. Experiments on the Libri2Mix dataset demonstrate that SEF-PNet significantly outperforms baseline models, achieving state-of-the-art PSE performance.
title SEF-PNet: Speaker Encoder-Free Personalized Speech Enhancement with Local and Global Contexts Aggregation
topic Audio and Speech Processing
Sound
url https://arxiv.org/abs/2501.11274