Unsupervised speaker adaptation using high confidence portion recognition results by multiple recognition systems.

Tomohiro Watanabe, Hiromitsu Nishizaki, Takehito Utsuro, Seiichi Nakagawa

Browse the full Interspeech paper archive.