Hi,
I’m trying to write some code to talk with HA in my native language.
Firstly, I have installed one Kaldi speech-to-text server in docker container, where I can extract detected speech in my native language.
Secondly, I’m sending this to google translate and receive back translation in English.
The Kaldi service is taking audio files as an input. How can I capture lovelace microphone feed so I could forward it to Kaldi input? Or are there other alternatives with keywords to activate recording?
Lastly, how can I sent the translated result to Almond?