Extracting deep bottleneck features for visual speech recognition.

Chao Sui, Roberto Togneri, Mohammed Bennamoun

Browse the full ICASSP paper archive.