Fig. 1From: Lightweight target speaker separation network based on joint trainingThe architecture of our proposed systemBack to article page