ARIEL has finally come alive (again)! Started off with importing the necessary modules and configuring input and output, making her sound as human as possible with whisper. Also integrated Ollama’s LLM into the voice assistant to deal with more complex queries.
Base features:
I’ve tested all these and they work pretty fine, the transcription is a bit dodgy though so might consider an ambient noise filter or something of the sort. Execution is also kinda slow, so might switch to a lower performance Ollama model.
Further feature ideas:
Excited to see how many of these survive the day!