Guardado en:
Detalles Bibliográficos
Autores principales: Song, Tianyu, Ta, Ton Viet, Thamwattana, Ngamta, Nomura, Hisako, Nguyen, Linh Thi Hoai
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:https://arxiv.org/abs/2605.12534
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913125594300416
author Song, Tianyu
Ta, Ton Viet
Thamwattana, Ngamta
Nomura, Hisako
Nguyen, Linh Thi Hoai
author_facet Song, Tianyu
Ta, Ton Viet
Thamwattana, Ngamta
Nomura, Hisako
Nguyen, Linh Thi Hoai
contents Most work in audio enhancement targets human speech, while bioacoustics is less studied due to noisy recordings and the distinct traits of animal sounds. To fill this gap, we adapt speech enhancement methods and build BioSEN, a model made for bioacoustic signals. BioSEN has three modules: a multi-scale dual-axis attention unit for time-frequency feature extraction, a bio-harmonic multi-scale enhancement unit for capturing harmonic structures, and an energy-adaptive gating connection unit that uses frequency weights to keep vocalizations from being removed as noise. Tests on three bioacoustic datasets show that BioSEN matches or exceeds state-of-the-art speech enhancement models while using far less computation. These results show BioSEN's strength for bioacoustic audio enhancement and its promise for biodiversity monitoring and conservation.
format Preprint
id arxiv_https___arxiv_org_abs_2605_12534
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle BioSEN: A Bio-acoustic Signal Enhancement Network for Animal Vocalizations
Song, Tianyu
Ta, Ton Viet
Thamwattana, Ngamta
Nomura, Hisako
Nguyen, Linh Thi Hoai
Sound
Machine Learning
Neurons and Cognition
Most work in audio enhancement targets human speech, while bioacoustics is less studied due to noisy recordings and the distinct traits of animal sounds. To fill this gap, we adapt speech enhancement methods and build BioSEN, a model made for bioacoustic signals. BioSEN has three modules: a multi-scale dual-axis attention unit for time-frequency feature extraction, a bio-harmonic multi-scale enhancement unit for capturing harmonic structures, and an energy-adaptive gating connection unit that uses frequency weights to keep vocalizations from being removed as noise. Tests on three bioacoustic datasets show that BioSEN matches or exceeds state-of-the-art speech enhancement models while using far less computation. These results show BioSEN's strength for bioacoustic audio enhancement and its promise for biodiversity monitoring and conservation.
title BioSEN: A Bio-acoustic Signal Enhancement Network for Animal Vocalizations
topic Sound
Machine Learning
Neurons and Cognition
url https://arxiv.org/abs/2605.12534