Fig. 1From: Non-parallel dictionary learning for voice conversion using non-negative Tucker decompositionBasic approach of NMF-based voice conversionBack to article page