Frequency-Weighted Training Losses for Phoneme-Level DNN-based Speech Enhancement

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Monir, Nasser-Eddine, Magron, Paul, Serizel, Romain
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866915355929083904
author Monir, Nasser-Eddine
Magron, Paul
Serizel, Romain
author_facet Monir, Nasser-Eddine
Magron, Paul
Serizel, Romain
contents Recent advances in deep learning have significantly improved multichannel speech enhancement algorithms, yet conventional training loss functions such as the scale-invariant signal-to-distortion ratio (SDR) may fail to preserve fine-grained spectral cues essential for phoneme intelligibility. In this work, we propose perceptually-informed variants of the SDR loss, formulated in the time-frequency domain and modulated by frequency-dependent weighting schemes. These weights are designed to emphasize time-frequency regions where speech is prominent or where the interfering noise is particularly strong. We investigate both fixed and adaptive strategies, including ANSI band-importance weights, spectral magnitude-based weighting, and dynamic weighting based on the relative amount of speech and noise. We train the FaSNet multichannel speech enhancement model using these various losses. Experimental results show that while standard metrics such as the SDR are only marginally improved, their perceptual frequency-weighted counterparts exhibit a more substantial improvement. Besides, spectral and phoneme-level analysis indicates better consonant reconstruction, which points to a better preservation of certain acoustic cues.
format Preprint
id arxiv_https___arxiv_org_abs_2506_18714
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Frequency-Weighted Training Losses for Phoneme-Level DNN-based Speech Enhancement
Monir, Nasser-Eddine
Magron, Paul
Serizel, Romain
Sound
Artificial Intelligence
Audio and Speech Processing
Recent advances in deep learning have significantly improved multichannel speech enhancement algorithms, yet conventional training loss functions such as the scale-invariant signal-to-distortion ratio (SDR) may fail to preserve fine-grained spectral cues essential for phoneme intelligibility. In this work, we propose perceptually-informed variants of the SDR loss, formulated in the time-frequency domain and modulated by frequency-dependent weighting schemes. These weights are designed to emphasize time-frequency regions where speech is prominent or where the interfering noise is particularly strong. We investigate both fixed and adaptive strategies, including ANSI band-importance weights, spectral magnitude-based weighting, and dynamic weighting based on the relative amount of speech and noise. We train the FaSNet multichannel speech enhancement model using these various losses. Experimental results show that while standard metrics such as the SDR are only marginally improved, their perceptual frequency-weighted counterparts exhibit a more substantial improvement. Besides, spectral and phoneme-level analysis indicates better consonant reconstruction, which points to a better preservation of certain acoustic cues.
title Frequency-Weighted Training Losses for Phoneme-Level DNN-based Speech Enhancement
topic Sound
Artificial Intelligence
Audio and Speech Processing
url https://arxiv.org/abs/2506.18714