Speaker-independent phone modeling based on speaker-dependent HMMs' composition and clustering.

Tetsuo Kosaka, Shoichi Matsunaga, Mikio Kuraoka

Browse the full ICASSP paper archive.