Separate Anything You Describe

Gardado en:

Detalles Bibliográficos
Publicado en:	arXiv.org (Dec 1, 2024), p. n/a
Autor Principal:	Liu, Xubo
Outros autores:	Kong, Qiuqiang, Zhao, Yan, Liu, Haohe, Yuan, Yi, Liu, Yuzhuo, Xia, Rui, Wang, Yuxuan, Plumbley, Mark D, Wang, Wenwu
Publicado:	Cornell University Library, arXiv.org
Materias:	Speech processing Source code Queries Scene analysis Musical instruments Separation Query languages Natural language
Acceso en liña:	Citation/Abstract Full text outside of ProQuest
Etiquetas:	Engadir etiqueta Sen Etiquetas, Sexa o primeiro en etiquetar este rexistro!

Descripción
Resumo:	Language-queried audio source separation (LASS) is a new paradigm for computational auditory scene analysis (CASA). LASS aims to separate a target sound from an audio mixture given a natural language query, which provides a natural and scalable interface for digital audio applications. Recent works on LASS, despite attaining promising separation performance on specific sources (e.g., musical instruments, limited classes of audio events), are unable to separate audio concepts in the open domain. In this work, we introduce AudioSep, a foundation model for open-domain audio source separation with natural language queries. We train AudioSep on large-scale multimodal datasets and extensively evaluate its capabilities on numerous tasks including audio event separation, musical instrument separation, and speech enhancement. AudioSep demonstrates strong separation performance and impressive zero-shot generalization ability using audio captions or text labels as queries, substantially outperforming previous audio-queried and language-queried sound separation models. For reproducibility of this work, we will release the source code, evaluation benchmark and pre-trained model at: https://github.com/Audio-AGI/AudioSep.
ISSN:	2331-8422
Fonte:	Engineering Database