How to generate speaker embedding for specific speaker?
#5
by
musharraf42
- opened
As mentioned in the README, the speaker embedding are hand-crafted, which makes it difficult to clone a specific speaker's voice.
Would you mind sharing the script used to generate the speaker embedding used in training? This way we can use the same approach to generate speaker embedding for arbitrary speakers.
Thank you for open-sourcing your work.
Best regards,
Musharraf