sanbancan team
VisionKnob is a local AI image-understanding assistant that combines Arduino UNO Q, Modulino hardware, Streamlit, and a Gemma vision model to let users physically select images and ask questions with private, on-device AI processing.