Fig. 2From: W2VC: WavLM representation based one-shot voice conversion with gradient reversal distillation and CTC supervisionThe architecture of the WavLM modelBack to article page