5 ms·

Hi all, author here. Besides the tech of Mimic 3 itself, I'm interested in training voices in as many (human) languages as possible. All it takes is one person

Hi all, author here. Besides the tech of Mimic 3 itself, I'm interested in training voices in as many (human) languages as possible. All it takes is one person willing to donate a dataset for everyone to benefit! ...well, that and a bunch of stuff with phonemes. But I'll do that part :)

dEnigma 4y ago
Can't you use the Mozilla Common Voice dataset for that?
- krisgesling 4y ago
  The Mozilla Common Voice dataset is awesome - however it's useful the opposite purpose - speech-to-text. This is because it is a lot of different people using a range of hardware, speaking similar phrases. For good text-to-speech you need 1 person speaking different phrases but very consistently. Here's an example dataset from Thorsten a German open voice enthusiast: https://openslr.org/95/ https://openslr.org/95/
  - dEnigma 4y ago
    Thanks for the explanation!
josephg 4y ago
How can people contribute? I'd be happy to sit in front of a microphone for awhile if I could use my own voice in a TTS engine!
- sampo 4y ago
  They want you to make good quality audio recordings of you speaking about 20 000 phrases. It could take 40 to 80 hours of speaking and recording, maximum 4 hours per day. https://github.com/MycroftAI/mimic-recording-studio https://github.com/MycroftAI/mimic-recording-studio https://mycroft.ai/contribute/ https://mycroft.ai/contribute/
  - synesthesiam 4y ago
    The amount of data depends on if there's a voice for the language already. If so, about 2 hours of data is usually good enough. Otherwise, 10-20 hours usually does it.
rjzzleep 4y ago
What does it take to add Chinese and Japanese to this? Surely it's a lot more than just training sets right? I have an android phone without access to google tts, so this might actually potentially be a nice alternative.
worthless-trash 4y ago
What kind of workload are we looking at, do you care for the Australian accent?
- krisgesling 4y ago
  Bloody oath we do!
  - krisgesling 4y ago
    Translation: "Yes" ... Hi from Darwin :D
wilsonjholmes 4y ago
Where could I donate my voice?

dEnigma 4y ago

Can't you use the Mozilla Common Voice dataset for that?

krisgesling 4y ago
The Mozilla Common Voice dataset is awesome - however it's useful the opposite purpose - speech-to-text. This is because it is a lot of different people using a range of hardware, speaking similar phrases. For good text-to-speech you need 1 person speaking different phrases but very consistently. Here's an example dataset from Thorsten a German open voice enthusiast: https://openslr.org/95/ https://openslr.org/95/
- dEnigma 4y ago
  Thanks for the explanation!

josephg 4y ago

How can people contribute? I'd be happy to sit in front of a microphone for awhile if I could use my own voice in a TTS engine!

sampo 4y ago
They want you to make good quality audio recordings of you speaking about 20 000 phrases. It could take 40 to 80 hours of speaking and recording, maximum 4 hours per day. https://github.com/MycroftAI/mimic-recording-studio https://github.com/MycroftAI/mimic-recording-studio https://mycroft.ai/contribute/ https://mycroft.ai/contribute/
- synesthesiam 4y ago
  The amount of data depends on if there's a voice for the language already. If so, about 2 hours of data is usually good enough. Otherwise, 10-20 hours usually does it.

rjzzleep 4y ago

What does it take to add Chinese and Japanese to this? Surely it's a lot more than just training sets right? I have an android phone without access to google tts, so this might actually potentially be a nice alternative.

worthless-trash 4y ago

What kind of workload are we looking at, do you care for the Australian accent?

krisgesling 4y ago
Bloody oath we do!
- krisgesling 4y ago
  Translation: "Yes" ... Hi from Darwin :D

wilsonjholmes 4y ago

Where could I donate my voice?