This project is an AI voice assistant that operates on a Raspberry Pi, capable of conducting multi-turn dialogues based on local LLM (Large Language Models) in an offline state, without the need for a wake-up phrase, ensuring complete local operation to protect your privacy. If equipped with a camera, it can even describe what it 'sees' or answer questions based on images.