3 Respostas
+ 3
Being on a Windows system might make things more tough simply because most of the fast language processing type libraries are going to have some C code with them and it's not unusual for that code to not be Windows friendly. You'll want to probably look into what the Mycroft project is doing since they've tackled a lot of this stuff. Bear in that their project isn't offline and that voice recognition and natural language processing is complicated and tends to be CPU intensive.
+ 3
Though my response might be late and perhaps half the solution, I highly recommend pythons SpeechRecognition library as well as pocketsphinx. As for text to speech a simple way would be to utilize win32com.client
for more info: https://stackoverflow.com/questions/1614059/how-to-make-JUMP_LINK__&&__python__&&__JUMP_LINK-speak
0
Hii