TTS and ASR

Hello all!

I have recently been looking into some TTS and speech recognition interfaces to borg into MRL. I may have just hit the jackpot!

I found two realistic-sounding TTS engines:

ResponsiveVoice.js and iSpeech

Both, unfortunately, require an internet connection but both are free. Again, unfortunately, neither are open source. ResponsiveVoice is written in JavaScript and iSpeech has an SDK for Java. iSpeech requires an API key while ResponsiveVoice does not. Both support a high number of voices and are both high quality.

iSpeech also includes a speech recognition interface, to provide an alternative to the Google stt engine.

Unfortuantely, I myself do not have time to implement these in MRL, I just wanted to throw these out there for anyone who's interested.

ShaunHolt's picture

Responsivevoice.js does seem

Responsivevoice.js does seem interesting and may be worth tinkering around with. Voices do sound pretty good. At a glance appears it may be quite simple to implement.

 

dom14's picture

You can use MicrosoftSpeech

You can use MicrosoftSpeech service.