Saved in:
Bibliographic Details
Main Authors: Tang, Beilong, Zeng, Bang, Li, Ming
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2504.07402
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915447837818880
author Tang, Beilong
Zeng, Bang
Li, Ming
author_facet Tang, Beilong
Zeng, Bang
Li, Ming
contents We propose LauraTSE, an Auto-Regressive Decoder-Only Language Model for Target Speaker Extraction built upon the LauraGPT backbone. LauraTSE employs a small-scale auto-regressive decoder-only language model that generates the initial layers of the target speech's discrete codec representations from the continuous embeddings of both the mixture and reference speech. These outputs serve as coarse-grained predictions. To refine them, a one-step encoder-only language model reconstructs the full codec representation by integrating information from both the mixture and the reference speech, adding fine-grained details. Experimental results show that our approach can achieve promising performance. Additionally, we conduct ablation studies to investigate the data scalability and the contribution of the encoder-only model.
format Preprint
id arxiv_https___arxiv_org_abs_2504_07402
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle LauraTSE: Target Speaker Extraction using Auto-Regressive Decoder-Only Language Models
Tang, Beilong
Zeng, Bang
Li, Ming
Machine Learning
Artificial Intelligence
We propose LauraTSE, an Auto-Regressive Decoder-Only Language Model for Target Speaker Extraction built upon the LauraGPT backbone. LauraTSE employs a small-scale auto-regressive decoder-only language model that generates the initial layers of the target speech's discrete codec representations from the continuous embeddings of both the mixture and reference speech. These outputs serve as coarse-grained predictions. To refine them, a one-step encoder-only language model reconstructs the full codec representation by integrating information from both the mixture and the reference speech, adding fine-grained details. Experimental results show that our approach can achieve promising performance. Additionally, we conduct ablation studies to investigate the data scalability and the contribution of the encoder-only model.
title LauraTSE: Target Speaker Extraction using Auto-Regressive Decoder-Only Language Models
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2504.07402