Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Yu, Chenxin, Ma, Hao, Li, Xu, Zhang, Xiao-Lei, Shao, Mingjie, Zhang, Chi, Li, Xuelong
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2510.13308
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909847483580416
author Yu, Chenxin
Ma, Hao
Li, Xu
Zhang, Xiao-Lei
Shao, Mingjie
Zhang, Chi
Li, Xuelong
author_facet Yu, Chenxin
Ma, Hao
Li, Xu
Zhang, Xiao-Lei
Shao, Mingjie
Zhang, Chi
Li, Xuelong
contents Query-based audio source extraction seeks to recover a target source from a mixture conditioned on a query. Existing approaches are largely confined to single-channel audio, leaving the spatial information in multi-channel recordings underexploited. We introduce a query-based spatial audio source extraction framework for recovering dry target signals from first-order ambisonics (FOA) mixtures. Our method accepts either an audio prompt or a text prompt as condition input, enabling flexible end-to-end extraction. The core of our proposed model lies in a tri-axial Transformer that jointly models temporal, frequency, and spatial channel dependencies. The model uses contrastive language-audio pretraining (CLAP) embeddings to enable unified audio-text conditioning via feature-wise linear modulation (FiLM). To eliminate costly annotations and improve generalization, we propose a label-free data pipeline that dynamically generates spatial mixtures and corresponding targets for training. The result of our experiment with high separation quality demonstrates the efficacy of multimodal conditioning and tri-axial modeling. This work establishes a new paradigm for high-fidelity spatial audio separation in immersive applications.
format Preprint
id arxiv_https___arxiv_org_abs_2510_13308
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Towards Multimodal Query-Based Spatial Audio Source Extraction
Yu, Chenxin
Ma, Hao
Li, Xu
Zhang, Xiao-Lei
Shao, Mingjie
Zhang, Chi
Li, Xuelong
Audio and Speech Processing
Query-based audio source extraction seeks to recover a target source from a mixture conditioned on a query. Existing approaches are largely confined to single-channel audio, leaving the spatial information in multi-channel recordings underexploited. We introduce a query-based spatial audio source extraction framework for recovering dry target signals from first-order ambisonics (FOA) mixtures. Our method accepts either an audio prompt or a text prompt as condition input, enabling flexible end-to-end extraction. The core of our proposed model lies in a tri-axial Transformer that jointly models temporal, frequency, and spatial channel dependencies. The model uses contrastive language-audio pretraining (CLAP) embeddings to enable unified audio-text conditioning via feature-wise linear modulation (FiLM). To eliminate costly annotations and improve generalization, we propose a label-free data pipeline that dynamically generates spatial mixtures and corresponding targets for training. The result of our experiment with high separation quality demonstrates the efficacy of multimodal conditioning and tri-axial modeling. This work establishes a new paradigm for high-fidelity spatial audio separation in immersive applications.
title Towards Multimodal Query-Based Spatial Audio Source Extraction
topic Audio and Speech Processing
url https://arxiv.org/abs/2510.13308