English

Universal Phone Recognition with a Multilingual Allophone System

Computation and Language 2020-02-28 v1 Sound Audio and Speech Processing

Abstract

Multilingual models can improve language processing, particularly for low resource situations, by sharing parameters across languages. Multilingual acoustic models, however, generally ignore the difference between phonemes (sounds that can support lexical contrasts in a particular language) and their corresponding phones (the sounds that are actually spoken, which are language independent). This can lead to performance degradation when combining a variety of training languages, as identically annotated phonemes can actually correspond to several different underlying phonetic realizations. In this work, we propose a joint model of both language-independent phone and language-dependent phoneme distributions. In multilingual ASR experiments over 11 languages, we find that this model improves testing performance by 2% phoneme error rate absolute in low-resource conditions. Additionally, because we are explicitly modeling language-independent phones, we can build a (nearly-)universal phone recognizer that, when combined with the PHOIBLE large, manually curated database of phone inventories, can be customized into 2,000 language dependent recognizers. Experiments on two low-resourced indigenous languages, Inuktitut and Tusom, show that our recognizer achieves phone accuracy improvements of more than 17%, moving a step closer to speech recognition for all languages in the world.

Keywords

Cite

@article{arxiv.2002.11800,
  title  = {Universal Phone Recognition with a Multilingual Allophone System},
  author = {Xinjian Li and Siddharth Dalmia and Juncheng Li and Matthew Lee and Patrick Littell and Jiali Yao and Antonios Anastasopoulos and David R. Mortensen and Graham Neubig and Alan W Black and Florian Metze},
  journal= {arXiv preprint arXiv:2002.11800},
  year   = {2020}
}

Comments

ICASSP 2020

R2 v1 2026-06-23T13:55:19.500Z