Ive managed to get my Atom Echo working as a voice assistant, and im using it with ChatGPT.
What I want to achieve is that it listens for follow up speech after the initial response, and if it it hears nothing, goes back to listening for a wake word after say 20 seconds.
I thought this would be pretty straightforward by modifying the code given for wakeword voice assist device to restart voice assistant in the ‘on_end’ part of the configuration but nomatter what I do it doesnt listen to anything else and instead I have to use the wakeword.
It seems other people want this a there is a feature request for exactly what I want. Has anyone managed to do anything like this with the currently available software? I dont really understand why adding ‘voice_assistant.start’ to the end of ‘on_end’ doesnt just…start the voice assistant as if the wake word had been spoken?