PURE Codec: Progressive Unfolding of Residual Entropy for Speech Codec Learning

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Shi, Jiatong, Wang, Haoran, Chen, William, Li, Chenda, Zhang, Wangyou, Tian, Jinchuan, Watanabe, Shinji
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918221564608512
author Shi, Jiatong
Wang, Haoran
Chen, William
Li, Chenda
Zhang, Wangyou
Tian, Jinchuan
Watanabe, Shinji
author_facet Shi, Jiatong
Wang, Haoran
Chen, William
Li, Chenda
Zhang, Wangyou
Tian, Jinchuan
Watanabe, Shinji
contents Neural speech codecs have achieved strong performance in low-bitrate compression, but residual vector quantization (RVQ) often suffers from unstable training and ineffective decomposition, limiting reconstruction quality and efficiency. We propose PURE Codec (Progressive Unfolding of Residual Entropy), a novel framework that guides multi-stage quantization using a pre-trained speech enhancement model. The first quantization stage reconstructs low-entropy, denoised speech embeddings, while subsequent stages encode residual high-entropy components. This design improves training stability significantly. Experiments demonstrate that PURE consistently outperforms conventional RVQ-based codecs in reconstruction and downstream speech language model-based text-to-speech, particularly under noisy training conditions.
format Preprint
id arxiv_https___arxiv_org_abs_2511_22687
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle PURE Codec: Progressive Unfolding of Residual Entropy for Speech Codec Learning
Shi, Jiatong
Wang, Haoran
Chen, William
Li, Chenda
Zhang, Wangyou
Tian, Jinchuan
Watanabe, Shinji
Sound
Audio and Speech Processing
Neural speech codecs have achieved strong performance in low-bitrate compression, but residual vector quantization (RVQ) often suffers from unstable training and ineffective decomposition, limiting reconstruction quality and efficiency. We propose PURE Codec (Progressive Unfolding of Residual Entropy), a novel framework that guides multi-stage quantization using a pre-trained speech enhancement model. The first quantization stage reconstructs low-entropy, denoised speech embeddings, while subsequent stages encode residual high-entropy components. This design improves training stability significantly. Experiments demonstrate that PURE consistently outperforms conventional RVQ-based codecs in reconstruction and downstream speech language model-based text-to-speech, particularly under noisy training conditions.
title PURE Codec: Progressive Unfolding of Residual Entropy for Speech Codec Learning
topic Sound
Audio and Speech Processing
url https://arxiv.org/abs/2511.22687