FurMe

FurMe won 1st overall at UofTHacks 11. It combines OpenCV face-tracking with an Arduino servo controller so the device physically rotates to keep the user centered, then layers emotion detection and LLMs for conversational context and Spotify playlist curation.

Contributions

• Closed-loop OpenCV face-tracking with Arduino servos to center the user in the camera viewport in real time.
• DeepFace emotion detection with Cohere and OpenAI for verbal responses and live sentiment-based Spotify playlists.
• IoT teleoperation dashboard in React and WebSockets for camera feed, keyboard servo control, and authenticated remote access.

Outcomes

• 1st overall at UofTHacks 11 among 600+ hackers.
• Working embodied AI demo with hardware, vision, and conversational layers.