Research Topics

  • Automatic development of high performance speech recognition systems based on evolutionary algorithms
  • Semi-supervised and unsupervised training of speech recognition systems
  • Kaldi recipe for high peformance Japanese speech recognition system (included in the Kaldi package)
  • Eye movement input based speech synthesis interface
  • Design and implementation of speech processing system based on functional programming
  • Automatic scoring of Japanese speaking test by applying speech recognition technology

Research introduction materials


Demos and software


Those who are interested in joining our lab

We welcome students who have passion for research on speech signal processing and machine learning. We host students from international graduate programs and research students as well as students from regular under graduate and graduate programs. If you are interested in joining our lab, please contact Takahiro Shinozaki. When you contact me for the application to the international programs, please appeal your motivation on research and skill on mathematics and programming attaching supporting materials, or I may not be able to respond. Information on programs for international study is available here.