Utatane Piko (ζ­Œζ‰‹ιŸ³γƒ”γ‚³) voice model for RVCv2

! Please credit me and don't use it for commercial purposes !

voice: Utatane Piko (VOCALOID 2)
dataset length: 1 hour
sample rate: 48 kHz
f0 method: RMVPE
pretrain: Default
Potentially multilingual

Utatane Piko (ζ­Œζ‰‹ιŸ³γƒ”γ‚³) is a Japanese VOCALOID developed and distributed by Ki/oon Music Inc., under Sony Music Entertainment Japan, and was released on December 8, 2010 for the VOCALOID2 engine. His voice is provided by the Japanese male singer PIKO.

This RVCv2 model was trained on a one-hour VOCALOID5-rendered dataset. It includes 10 minutes of basic sounds, like the original voicebank, plus specific syllables such as [a h_0] or [4 a 4 a 4 a 4 a 4 a]+[4 a] (β€œTrilled R”), and VOCALOID5-generated whisper. The remaining 50 minutes of the dataset consist of rendered singing, based on 19 VSQx files. The dataset contains singing ranging from C2 to C7 (which is C1 to C6 in Vocaloid).

model demo songs:
FLAVOR FOLEY - weathergirl: https://youtu.be/1_qqaY9Job8?si=D6yq2S9aXvXPtUxy
The Living Tombstone - I Got No Time: https://youtu.be/BbESBqj-T5I?si=VMiOYANTWEclADyC

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support