A new speech coding model based on a least-squares sinusoidal representation.

E. Bryan George, Mark J. T. Smith

Browse the full ICASSP paper archive.