Viseme-dependent weight optimization for CHMM-based audio-visual speech recognition.

Alexey Karpov, Andrey Ronzhin, Konstantin Markov, Milos Zelezn

Browse the full Interspeech paper archive.