Using twin-HMM-based audio-visual speech enhancement as a front-end for robust audio-visual speech recognition.

Ahmed Hussen Abdelaziz, Steffen Zeiler, Dorothea Kolossa

Browse the full Interspeech paper archive.