The ASR Post-Processor Performance Challenges of BackTranScription (BTS): Data-Centric and Model-Centric Approaches

Chanjun Park, Jaehyung Seo, Seolhwa Lee, Chanhee Lee, Heuiseok Lim

Research output: Contribution to journalArticlepeer-review

Abstract

Training an automatic speech recognition (ASR) post-processor based on sequence-to-sequence (S2S) requires a parallel pair (e.g., speech recognition result and human post-edited sentence) to construct the dataset, which demands a great amount of human labor. BackTransScription (BTS) proposes a data-building method to mitigate the limitations of the existing S2S based ASR post-processors, which can automatically generate vast amounts of training datasets, reducing time and cost in data construction. Despite the emergence of this novel approach, the BTS-based ASR post-processor still has research challenges and is mostly untested in diverse approaches. In this study, we highlight these challenges through detailed experiments by analyzing the data-centric approach (i.e., controlling the amount of data without model alteration) and the model-centric approach (i.e., model modification). In other words, we attempt to point out problems with the current trend of research pursuing a model-centric approach and alert against ignoring the importance of the data. Our experiment results show that the data-centric approach outperformed the model-centric approach by +11.69, +17.64, and +19.02 in the F1-score, BLEU, and GLEU tests.

Original languageEnglish
Article number3618
JournalMathematics
Volume10
Issue number19
DOIs
Publication statusPublished - 2022 Oct

Keywords

  • automatic speech recognition
  • backtranscription
  • data-centric
  • machine translation
  • model-centric
  • post-processor

ASJC Scopus subject areas

  • Computer Science (miscellaneous)
  • Mathematics(all)
  • Engineering (miscellaneous)

Fingerprint

Dive into the research topics of 'The ASR Post-Processor Performance Challenges of BackTranScription (BTS): Data-Centric and Model-Centric Approaches'. Together they form a unique fingerprint.

Cite this