On the Subject of FaceFinder
• Developed a Progressive Web App (PWA) using React.js and Python Flask API that scans faces and retrieves matching images and videos using InsightFace.
• Implemented vector embeddings with cosine similarity to enable accurate facial matching and display confidence percentages for each result.
• Designed a feature to analyze videos and detect timestamps where the identified person appears, improving search efficiency.
• Built a secure system with user authentication and an admin panel to manage media (add/remove images and videos).
• Created a personalized user gallery that stores previously matched results, ensuring users only see new and relevant data in future scans.
• Deployed the application using Netlify, Hugging Face, and Supabase, with source code managed on GitHub.
Project Genre
Technologies Used