search

lowerquality / gentle

gentle forced aligner
https://lowerquality.com/gentle/
MIT License
1.46k stars 296 forks source link

IPA Transcription #70

Open jfg2131 opened 8 years ago

jfg2131 commented 8 years ago

Is the English-language transcription intentional? The International Phonetic Alphabet would be generalizable and more precise.

maxhawkins commented 8 years ago

Thank you for the message! We use ARPAbet for phonetic transcriptions because the speech model we use is based on the CMU Pronouncing Dictionary.

I added a page to the wiki with some more information about our pre-packaged model. It is not yet possible to use another model with Gentle. However, this is something we'd like to do.

There are versions of the CMU Pronouncing Dictioanry that use IPA. In the future we may build a model that uses this alternate dictionary.

See ARPAbet wikipedia page for more information about how the symbols map to IPA.

jfg2131 commented 8 years ago

Thanks for the quick response!

Jake

On Fri, Jan 29, 2016 at 12:54 PM, Max Hawkins notifications@github.com wrote:

Thank you for the message! We use ARPAbet https://en.wikipedia.org/wiki/Arpabet for phonetic transcriptions because the speech model we use is based on the CMU Pronouncing Dictionary http://www.speech.cs.cmu.edu/cgi-bin/cmudict.

I added a page to the wiki https://github.com/lowerquality/gentle/wiki/Model with some more information about our pre-packaged model. It is not yet possible to use another model with Gentle. However, this is something we'd like to do.

There are versions of the CMU Pronouncing Dictioanry that use IPA. In the future we may build a model that uses this alternate dictionary.

See ARPAbet wikipedia page https://en.wikipedia.org/wiki/Arpabet for more information about how the symbols map to IPA.

— Reply to this email directly or view it on GitHub https://github.com/lowerquality/gentle/issues/70#issuecomment-176965466.