Paper

CFVC: Conditional Filtering for Controllable Voice Conversion

Kou Tanaka, Takuhiro Kaneko, Hirokazu Kameoka, Shogo Seki

Audio samples for speaking rate control

Source speaker: mho (male) mht (male) mmy (male) msh (male) mtk (male) myi (male) fkn (female) fks (female) ftk (female) fym (female) --> Target speaker: mho (male) mht (male) mmy (male) msh (male) mtk (male) myi (male) fkn (female) fks (female) ftk (female) fym (female)
(Supported: Safari, Chrome, FireFox, Opera)

Re-synthsized		Baseline	Proposed
Source	Target		0.6	0.8	1.0	1.2	1.4	Swap

Bonus samples for pitch control

Source speaker: mho (male) mht (male) mmy (male) msh (male) mtk (male) myi (male) fkn (female) fks (female) ftk (female) fym (female) --> Target speaker: mho (male) mht (male) mmy (male) msh (male) mtk (male) myi (male) fkn (female) fks (female) ftk (female) fym (female)
(Supported: Safari, Chrome, FireFox, Opera)

Re-synthsized		Baseline	Proposed
Source	Target		0.6	0.8	1.0	1.2	1.4	Swap