This entire idea originated from the SIH 2025 Problem Statement for the NSG I've been working on. I was already deep into designing a complex, multi-stage system to solve their challenges: tracking a PoI across non-overlapping cameras , handling disguises using gait and biometrics , and running threat analysis. When the "Hackdays for AI" event came up, I realized I already had a powerful, well-researched concept. I'm essentially adapting my comprehensive solution for the NSG—which covers everything from YOLOv8 for detection , BoT-SORT for tracking , FastReID for cross-camera identification , and Isolation Forest for anomaly detection —to fit the hackathon. It's a robust, real-world solution ready to be built. Inspiration

Built With

Share this project:

Updates