Incorporating Lip Features into Audio-Visual Multi-Speaker DOA Estimation by Gated Fusion.

Ya Jiang, Hang Chen, Jun Du, Qing Wang, Chin-Hui Lee

Browse the full ICASSP paper archive.