1 paper
Pengjie Shen, Kangrui Chen, Shulin He +5
We propose listen to extract (LExt), a highly-effective while extremely-simple algorithm for monaural target speaker extraction (TSE). Given an enrollment utterance of a target spe…