Target Speaker Extraction by Directly Exploiting Contextual Information in the Time-Frequency Domain.

Xue Yang, Changchun Bao, Jing Zhou, Xianhong Chen

Browse the full ICASSP paper archive.