PHRASED: Phrase Dictionary Biasing for Speech Translation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Peidong, Xue, Jian, Zhao, Rui, Chen, Junkun, Subramanian, Aswin Shanmugam, Li, Jinyu
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912424418869248
author Wang, Peidong
Xue, Jian
Zhao, Rui
Chen, Junkun
Subramanian, Aswin Shanmugam
Li, Jinyu
author_facet Wang, Peidong
Xue, Jian
Zhao, Rui
Chen, Junkun
Subramanian, Aswin Shanmugam
Li, Jinyu
contents Phrases are essential to understand the core concepts in conversations. However, due to their rare occurrence in training data, correct translation of phrases is challenging in speech translation tasks. In this paper, we propose a phrase dictionary biasing method to leverage pairs of phrases mapping from the source language to the target language. We apply the phrase dictionary biasing method to two types of widely adopted models, a transducer-based streaming speech translation model and a multimodal large language model. Experimental results show that the phrase dictionary biasing method outperforms phrase list biasing by 21% relatively for the streaming speech translation model. In addition, phrase dictionary biasing enables multimodal large language models to use external phrase information, achieving 85% relative improvement in phrase recall.
format Preprint
id arxiv_https___arxiv_org_abs_2506_09175
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle PHRASED: Phrase Dictionary Biasing for Speech Translation
Wang, Peidong
Xue, Jian
Zhao, Rui
Chen, Junkun
Subramanian, Aswin Shanmugam
Li, Jinyu
Computation and Language
Artificial Intelligence
Sound
Audio and Speech Processing
Phrases are essential to understand the core concepts in conversations. However, due to their rare occurrence in training data, correct translation of phrases is challenging in speech translation tasks. In this paper, we propose a phrase dictionary biasing method to leverage pairs of phrases mapping from the source language to the target language. We apply the phrase dictionary biasing method to two types of widely adopted models, a transducer-based streaming speech translation model and a multimodal large language model. Experimental results show that the phrase dictionary biasing method outperforms phrase list biasing by 21% relatively for the streaming speech translation model. In addition, phrase dictionary biasing enables multimodal large language models to use external phrase information, achieving 85% relative improvement in phrase recall.
title PHRASED: Phrase Dictionary Biasing for Speech Translation
topic Computation and Language
Artificial Intelligence
Sound
Audio and Speech Processing
url https://arxiv.org/abs/2506.09175