TTS-Transducer: End-to-End Speech Synthesis with Neural Transducer

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F26%3A2EC93SW8" target="_blank" >RIV/00216208:11320/26:2EC93SW8 - isvavai.cz</a>

Výsledek na webu

<a href="https://ieeexplore.ieee.org/abstract/document/10890256" target="_blank" >https://ieeexplore.ieee.org/abstract/document/10890256</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/ICASSP49660.2025.10890256" target="_blank" >10.1109/ICASSP49660.2025.10890256</a>

Jazyk výsledku

angličtina

Název v původním jazyce

TTS-Transducer: End-to-End Speech Synthesis with Neural Transducer

Popis výsledku v původním jazyce

This work introduces TTS-Transducer – a novel architecture for text-to-speech, leveraging the strengths of audio codec models and neural transducers. Transducers, renowned for their superior quality and robustness in speech recognition, are employed to learn monotonic alignments and allow for avoiding using explicit duration predictors. Neural audio codecs efficiently compress audio into discrete codes, revealing the possibility of applying text modeling approaches to speech generation. However, the complexity of predicting multiple tokens per frame from several codebooks, as necessitated by audio codec models with residual quantizers, poses a significant challenge. The proposed system first uses a transducer architecture to learn monotonic alignments between tokenized text and speech codec tokens for the first codebook. Next, a non-autoregressive Transformer predicts the remaining codes using the alignment extracted from transducer loss. The proposed system is trained end-to-end. We show that TTS-Transducer is a competitive and robust alternative to contemporary TTS systems1.

Název v anglickém jazyce

TTS-Transducer: End-to-End Speech Synthesis with Neural Transducer

Popis výsledku anglicky

This work introduces TTS-Transducer – a novel architecture for text-to-speech, leveraging the strengths of audio codec models and neural transducers. Transducers, renowned for their superior quality and robustness in speech recognition, are employed to learn monotonic alignments and allow for avoiding using explicit duration predictors. Neural audio codecs efficiently compress audio into discrete codes, revealing the possibility of applying text modeling approaches to speech generation. However, the complexity of predicting multiple tokens per frame from several codebooks, as necessitated by audio codec models with residual quantizers, poses a significant challenge. The proposed system first uses a transducer architecture to learn monotonic alignments between tokenized text and speech codec tokens for the first codebook. Next, a non-autoregressive Transformer predicts the remaining codes using the alignment extracted from transducer loss. The proposed system is trained end-to-end. We show that TTS-Transducer is a competitive and robust alternative to contemporary TTS systems1.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

—

Rok uplatnění

2025

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

ISBN

—

ISSN

15206149

e-ISSN

—

Počet stran výsledku

5

Strana od-do

1-5

Název nakladatele

—

Místo vydání

—

Místo konání akce

Hyderabad

Datum konání akce

1. 1. 2026

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—