Noisy speech enhancement based on prior knowledge about spectral envelope and harmonic structure.

Takuya Yoshioka, Tomohiro Nakatani, Hiroshi G. Okuno

Browse the full ICASSP paper archive.