Won 1st place for outstanding teamwork at the prestigious AI+ Global Health Hackathon in San Francisco, competing against teams from top tech companies and universities.
Project: LumiWalk
Developed an AI-powered visual assistant platform to help visually impaired individuals navigate their environment safely through real-time audio feedback.
Key Features Developed
- Real-time obstacle detection using YOLOv8
- Spatial audio feedback system for directional guidance
- Apple Vision Pro integration for depth sensing
- Natural language scene descriptions using VLMs
- Voice command interface for hands-free operation
Team Contribution
- Led computer vision pipeline development
- Integrated multiple AI models into unified system
- Coordinated 4-person team through 24-hour sprint