Dual-Path Cross-Modal Attention for Better Audio-Visual Speech Extraction.

Zhongweiyang Xu, Xulin Fan, Mark Hasegawa-Johnson

Browse the full ICASSP paper archive.