How about numbers or complex/unknown words? I would love to proxy everything that isn't known AFTER the offline trigger word to Google voice recognition or something. Is that possible?
Obviously, you can just grab the audio stream after the phrase is detected and route it to whatever you like. Google ASR or even a local one running on the device!
I have literally 3 projects, one a current big that depend or would heavily benefit from this. Can't wait to give it a try!