Some Problems on Automatic Speech Recognition

  • NAKAGAWA Seiichi
    Department of Information and Computer Sciences, Toyohashi University of Technology

Bibliographic Information

Other Title
  • 音声認識研究の課題

Search this article

Description

In this paper, we discuss some problems on automatic speech recognition. First,we state that the current art of speech recognition technology is inferior to the ability of human beings, especially, on phoneme recognition(perception)performance. Therefore, we point out the importance of acoustic models. Next, we describe a blance a between an acoustic model and a language model from a viewpoint of information theory. Finally, we focus on research trends on statistics-based language model and we describe the limits and future works.

Journal

References(25)*help

See more

Details 詳細情報について

  • CRID
    1570009752533214720
  • NII Article ID
    110003278509
  • NII Book ID
    AN10091225
  • Text Lang
    ja
  • Data Source
    • CiNii Articles

Report a problem

Back to top