Demonstration of single channel speech dereverberation

     I made a recording of my voice with a monoral voice recorder and tried a dereverberation. During the recording, voice recorder was held with my hand and kept relatively still.

  1. This sound clip contains recorded speech.
  2. This sound clip contains  dereverberated speech processed by the "multi-step linear prediction based dereverberation". (Ephraim's spectral subtraction was employed as a pre-processing to suppress background noise.)


1. K. Kinoshita, T. Nakatani and M. Miyoshi, `` Spectral subtraction steered by multi-step forward linear prediction  for single channel speech dereverberation,'' Proc. Of  International Conference on Acoustics, Speech, and Signal Processing(ICASSP), I, pp.817-820, 2006 [Presented poster]

2. K. Kinoshita, M. Delcroix, T. Nakatani and M. Miyoshi, “A linear prediction-based microphone array for speech dereverberation in a realistic sound field,’’ Audio Engineering Society (AES) 13th Regional Convention, Tokyo, 2007 [Presented poster]

Keisuke Kinoshita at