End-to-End multi-talker audio-visual ASR using an active speaker attention module.

Richard Rose, Olivier Siohan

Browse the full Interspeech paper archive.