Listening skills from the perspective of speech recognition

DOI

Bibliographic Information

Other Title
  • 音声認識から見たリスニング能力
  • Language model is what you need
  • 文法と語彙の重要性

Abstract

This paper explores the elements necessary for listening by utilizing a non-end-to-end speech recognition model. While keeping the acoustic model fixed, which corresponds to good hearing in the case of humans, the dictionary and grammar models are exchanged for verification. The article highlights the importance of vocabulary and grammar in listening instruction, noting that without vocabulary and grammar, there is a significant difference in recognition accuracy.

Journal

Keywords

Details 詳細情報について

  • CRID
    1390862345537906304
  • DOI
    10.11480/kyoyobukiyo.54.0_81
  • ISSN
    2433359X
    03863492
  • Text Lang
    en
  • Data Source
    • JaLC
  • Abstract License Flag
    Disallowed

Report a problem

Back to top