Improving audio-visual speech recognition using deep neural networks with dynamic stream reliability estimates.

Hendrik Meutzner, Ning Ma, Robert M. Nickel, Christopher Schymura, Dorothea Kolossa

Browse the full ICASSP paper archive.