3

I am working on a project which needs to convert a given speech (in an audio file,only speech no music and other stuffs) into text and later analyze that text. It would be a great help if I could able to find a open source tool which would help me in this regard.

Thanks in advance !

3
  • 1
    Only if you already know how to do it. Things can be put together from parts available; but you will have to do the actual speech-to-spelling yourself, because it's fiendishly complex if you don't use phonology as an interface. If you do, it's only terribly complex. And companies who've done the work keep it very, very close. Probly your best bet is to simply buy the best commercial software.
    – jlawler
    Jan 21 '14 at 16:43
  • @jlawler Would you mind listing good speech-to-IPA and vice versa projects you may know about? I for one would be willing to spend the time and effort to put those pieces together.
    – Joe Pineda
    Jan 24 '14 at 16:48
  • Dragon is a good commercial speech-to-text project, but it doesn't do IPA at all. Neither does Siri on iphones. You'd hafta add a text-to-IPA module, and that means you'd hafta pick a dialect to use. RP, American, Oz, NZ, S.Africa, Jamaican, Indian, Chinese, many more. All different phonemically, and therefore even more different phonetically in IPA; but not different in text. Which is one reason Dragon and Siri stop there.
    – jlawler
    Jan 24 '14 at 17:47
1

CMU Sphinx is a good library: link

1

Regarding the performances, see: How does Kaldi compare with Mozilla DeepSpeech in terms of speech recognition accuracy?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.