Fig. 1From: Multimodal voice conversion based on non-negative matrix factorizationBasic approach of NMF-based voice conversionBack to article page