Compensation of speaker directivity in speech recognition using HMM composition.

Franck Giron, Yasuhiro Minami, Masashi Tanaka, Ken'ichi Furuya

Browse the full ICASSP paper archive.