A voice separation system based on phoneme segregation
Ni You-Seng, Jian Chen, Chen Hong-Bing · 1992
Presents a new voice separation system for improving the intelligibility of speech that has been interfered by another talker's voice. The system is based on a sinusiodal model of phoneme and a phoneme segregation algorithm, both are developed. The primary results of separation experiments performed on natural voice are encouraging. The system presented could be used as a preprocessor for speech recognition or as a part of a digital signal processing hearing aid for those hearing-impaired listeners.>