A stream-weight optimization method for audio-visual speech recognition using multi-stream HMMs.

Satoshi Tamura, Koji Iwano, Sadaoki Furui

Browse the full ICASSP paper archive.