1 paper
Ke Zhang, Junjie Li, Shuai Wang +4
Target speaker extraction (TSE) relies on a reference cue of the target to extract the target speech from a speech mixture. While a speaker embedding is commonly used as the refere…