Feature to take audio/voice input
bhawesh96 opened this issue · 1 comments
Some Mitra might want to simply input a recording of His speech or actually speak out his/her (mitra's) mann ki baat xD
Adding pyaudio and some NLP with the existing lexing and parsing should get this working!
Acche din aa gaye
Hi @bhawesh96
Thanks for the suggestion but ASR for Hindi might be a tough task to solve. Certainly there are pre-trained models out there but fine-tuning them to the scope of this project might still require some time. It's not something that we plan on doing for now as it might require a significant amount of time to get speech recognition on Hindi working that's fine-tuned for this project.
Closing this thread as we don't plan to add for now. Although we would be happy to accept any PR in this regard.