This project combines facial landmark detection and voice interaction to create an expressive chatbot. The face tracking module uses OpenCV and Dlib to detect facial expressions and drive a Live2D animated model. The AI module handles voice-to-text transcription via Iflytek IAT, processes responses through the Turing chatbot API, and synthesizes speech output via Iflytek TTS. The result is a voice chatbot that responds conversationally while visually mirroring the user’s expressions.