Research Methods

This multi-disciplinary programme involves a variety of methodologies:
  • Psychoacoustic experimentation, including the use of virtual sound sources.
  • Computer modeling of auditory processing and speech production, including neural oscillator models, room reverberation models and speech production models.
  • Evaluation of these models by, for instance, speech recognition or speech enhancement studies.
  • Automatic Speech Recognition (ASR) techniques,

  • ASR evaluation.