Detective Conan Child's Voice / Shota Voice RVC model
Inference Comparison
After Inference
Model Introduction
Last release Haibara Aitai / Japanese dual-duo RVC children's voice model received a pretty good response!
Today, I'm bringing you the Detective series version of Conan. For the original Conan dataset, we used Taiwan voice acting + Japanese dub bilingual training datasets. The Taiwan voice dataset includes data from episodes 1 to 600, plus data from the Japanese dub movie set. Therefore, the model's voice can be either childish or boyish.
Due to the special nature of this model, using the original voice can be somewhat troublesome. So if your original tone differs from Miaoin's or Conan's tone during use, try to imitate Conan's tone or increase or decrease the pitch to achieve the desired effect. This model works even better for female voices, but for male voices, you need to raise the pitch! ~
Below are several comparisons of the original reasoning before and after results.
Not reasoning (female voice).
<!--[if lt IE 9]>document.createElement('audio');<![endif]-->After reasoning
Not reasoning (male voice).
After reasoning
I set the pitch cap to position 10 and adjust it according to your own effect. If you can speak a strong stage accent, that's even better.
Japanese expression
Parameters
Tone change: 0
Algorithm: rmvpe
Singing: Not supported
Languages: Chinese/Japanese
Training: 200
Sampling rate: 44K
Training log
![Image[1]-Miaoyin-RVC Sound Model Workshop](https://apis.klrvc.com/wp-content/uploads/2024/07/20240721142200458-WX20240721-135508-300x173.png)
User 93954589
Version Details
Recommended Parameters
Related Models
Keywords
Comments
No comments yet

