1 paper
Moshe Mandel, Shlomo E. Chazan
We present a voice conversion (VC) framework that utilizes K-Nearest Neighbors (KNN) retrieval over WavLM representations to align non-parallel source and target speech, constructi…