Maybe a collaboration would be interesting, see this link
1 Like
Hi and welcome to the community!
Moving to #tts category since the project seems to be more related with it.
@kdavis it seems this people are collecting a lot of Spanish voices for TTS, it might be interesting to reach out for collaboration on the dataset license?
Cheers.
Do you more details? The link doesn’t say much.
Let me ask the reported who covered the site when it was up, he might have a contact at the University.
Thats exactly what I was going to suggest. The site is apparently down due to huge demand, as can be read here: https://aholab.ehu.eus/aholab/
Ok, I’ve emailed the people leading the project to get more info. I’ll report back.
The reply I got is that they would like their data to be available for open use for investigation purposes, but in this specific project they didn’t get the permissions to share any of the dataset they are collecting.
Also, they mention that their dataset is not useful for general TTS, there are just a few sentences per speaker and always the same ones. They use these “HTS voices” only for adaptation.
Cheers.
1 Like