Non-linear representations, sensor reliability estimation and context-dependent fusion in the audiovisual recognition of speech in noise.

Pascal Teissier, Jean-Luc Schwartz, Anne Gurin-Dugu

Browse the full Interspeech paper archive.