English

Segmenting Subtitles for Correcting ASR Segmentation Errors

Computation and Language 2021-04-19 v1

Abstract

Typical ASR systems segment the input audio into utterances using purely acoustic information, which may not resemble the sentence-like units that are expected by conventional machine translation (MT) systems for Spoken Language Translation. In this work, we propose a model for correcting the acoustic segmentation of ASR models for low-resource languages to improve performance on downstream tasks. We propose the use of subtitles as a proxy dataset for correcting ASR acoustic segmentation, creating synthetic acoustic utterances by modeling common error modes. We train a neural tagging model for correcting ASR acoustic segmentation and show that it improves downstream performance on MT and audio-document cross-language information retrieval (CLIR).

Keywords

Cite

@article{arxiv.2104.07868,
  title  = {Segmenting Subtitles for Correcting ASR Segmentation Errors},
  author = {David Wan and Chris Kedzie and Faisal Ladhak and Elsbeth Turcan and Petra Galuščáková and Elena Zotkina and Zhengping Jiang and Peter Bell and Kathleen McKeown},
  journal= {arXiv preprint arXiv:2104.07868},
  year   = {2021}
}
R2 v1 2026-06-24T01:13:42.874Z