Noise robust front-end processing with voice activity detection based on periodic to aperiodic component ratio.

Kentaro Ishizuka, Tomohiro Nakatani, Masakiyo Fujimoto, Noboru Miyazaki

Browse the full Interspeech paper archive.