Can I replace the espeak-ng phonemizer with neural g2p
I was wondering as the piper tts works with rule based phonemizer can I replace it with neural g2p, so it is more robust and adaptable to Any
accents
waitting for your work.i noticed the same issu,some missed phoneme can be fix by espeak-ng.https://github.com/rhasspy/piper/issues/344#issuecomment-1890844487
I see many tts todays like orpheus, dia tts and sesame ai tts, and maybe eleven labs. the all have language model in btw the architecture , that helps drive the natural prosody, laughs, sighs etc. I was wondering can piper tts be engineered in that way
I was thinking the same of adding more phonemizers to it. Technically eSpeakNG is optional, but because of... aha, license conflict detected! Maybe that phonemizer you've requested or some others I don't know?