Speaker adaptive phoneme recognition by multi-level clustering based on mutual information criterion.

Katsuhiko Shirai, Naoki Hosaka, Eiichiro Kitagawa

Browse the full ICASSP paper archive.