X-SEPFORMER: End-To-End Speaker Extraction Network with Explicit Optimization on Speaker Confusion.

Kai Liu, Ziqing Du, Xucheng Wan, Huan Zhou

Browse the full ICASSP paper archive.