search

dmort27 / epitran

A tool for transcribing orthographic text as IPA (International Phonetic Alphabet)
MIT License
625 stars 120 forks source link

The American IPA output #161

Open JohnHerry opened 10 months ago

JohnHerry commented 10 months ago

Thanks for the good job, But it seems that the IPA from flite are tone of UK, is there any solution for American type of IPA transcription?

dmort27 commented 10 months ago

Presently, only Standard American English is supported (since we do not have a good open source pronouncing dictionary for RP and this would be necessary for training a redistributable model for G2P.

JohnHerry commented 10 months ago

Thanks for the quick reply, And I have another question. there seems no stress syllable in the output IPA. the stress symbol should be an important part in the TTS input phonemes. is there any plan to support it?

kalvinchang commented 10 months ago

Consider the CMU Pronouncing Dictionary.

On Mon, Oct 16, 2023 at 7:40 PM JohnHerry @.***> wrote:

Thanks for the quick reply, And I have another question. there seems no stress syllable in the output IPA. the stress symbol should be an important part in the TTS input phonemes. is there any plan to support it?

— Reply to this email directly, view it on GitHub https://github.com/dmort27/epitran/issues/161#issuecomment-1765564523, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEZUVP2TNNWIG2WATSPAIQ3X7XVYHAVCNFSM6AAAAAA6BYCQ62VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONRVGU3DINJSGM . You are receiving this because you are subscribed to this thread.Message ID: @.***>

JohnHerry commented 10 months ago

Consider the CMU Pronouncing Dictionary. … On Mon, Oct 16, 2023 at 7:40 PM JohnHerry @.> wrote: Thanks for the quick reply, And I have another question. there seems no stress syllable in the output IPA. the stress symbol should be an important part in the TTS input phonemes. is there any plan to support it? — Reply to this email directly, view it on GitHub <#161 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEZUVP2TNNWIG2WATSPAIQ3X7XVYHAVCNFSM6AAAAAA6BYCQ62VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONRVGU3DINJSGM . You are receiving this because you are subscribed to this thread.Message ID: @.>

Thank you for the help. But can the dict based g2p distinguish polyphonic words?

kalvinchang commented 10 months ago

CMU Pronouncing Dictionary lists multiple pronunciations in a few cases but unfortunately, not for most words (iirc)

JohnHerry commented 10 months ago

@kalvinchang Thanks, Do you have any better resolution? We had tried to use espeak-ng, but it is hard to compile on my Centos7 system.