SeamlessExpressiveLM: Speech Language Model for Expressive Speech-to-Speech Translation with Chain-of-Thought

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Gong, Hongyu, Veluri, Bandhav
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917679711911936
author Gong, Hongyu
Veluri, Bandhav
author_facet Gong, Hongyu
Veluri, Bandhav
contents Expressive speech-to-speech translation (S2ST) is a key research topic in seamless communication, which focuses on the preservation of semantics and speaker vocal style in translated speech. Early works synthesized speaker style aligned speech in order to directly learn the mapping from speech to target speech spectrogram. Without reliance on style aligned data, recent studies leverage the advances of language modeling (LM) and build cascaded LMs on semantic and acoustic tokens. This work proposes SeamlessExpressiveLM, a single speech language model for expressive S2ST. We decompose the complex source-to-target speech mapping into intermediate generation steps with chain-of-thought prompting. The model is first guided to translate target semantic content and then transfer the speaker style to multi-stream acoustic units. Evaluated on Spanish-to-English and Hungarian-to-English translations, SeamlessExpressiveLM outperforms cascaded LMs in both semantic quality and style transfer, meanwhile achieving better parameter efficiency.
format Preprint
id arxiv_https___arxiv_org_abs_2405_20410
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle SeamlessExpressiveLM: Speech Language Model for Expressive Speech-to-Speech Translation with Chain-of-Thought
Gong, Hongyu
Veluri, Bandhav
Computation and Language
Artificial Intelligence
Sound
Audio and Speech Processing
Expressive speech-to-speech translation (S2ST) is a key research topic in seamless communication, which focuses on the preservation of semantics and speaker vocal style in translated speech. Early works synthesized speaker style aligned speech in order to directly learn the mapping from speech to target speech spectrogram. Without reliance on style aligned data, recent studies leverage the advances of language modeling (LM) and build cascaded LMs on semantic and acoustic tokens. This work proposes SeamlessExpressiveLM, a single speech language model for expressive S2ST. We decompose the complex source-to-target speech mapping into intermediate generation steps with chain-of-thought prompting. The model is first guided to translate target semantic content and then transfer the speaker style to multi-stream acoustic units. Evaluated on Spanish-to-English and Hungarian-to-English translations, SeamlessExpressiveLM outperforms cascaded LMs in both semantic quality and style transfer, meanwhile achieving better parameter efficiency.
title SeamlessExpressiveLM: Speech Language Model for Expressive Speech-to-Speech Translation with Chain-of-Thought
topic Computation and Language
Artificial Intelligence
Sound
Audio and Speech Processing
url https://arxiv.org/abs/2405.20410