This growing (hopefully) list of examples demonstrates some of the ideas I've been toying with over the past several months regarding the perceptual results when speech sounds are modified in certain ways. The examples may seem somewhat disconnected, as I'm only just beginning to formulate my own ideas on how to increase speech recognition performance. I hope, however, that at some point they'll coalesce into something useful.

Where's the information contained in a speech signal

  • Scaling speech by appropriately placed Gaussians I.
  • Scaling speech by appropriately placed Gaussians II.
  • Auditory system resolution

  • The effect of linear frequency shifting on speech intelligibility.
  • Please send any questions, comments, or corrections to me

    Return to my Home page.