3rd YearAdvancedAI/ML 10-12 weeks
Smart Attendance System with Face Recognition
An automated attendance marking system using face recognition — students register their face; the system automatically marks attendance when they appear in the classroom camera feed.
PythonOpenCVface_recognitionFlaskPostgreSQLReact
Find GitHub repos & source code →How to build it — step by step
- 1Face Enrollment: Capture 5-10 photos per student from different angles; extract 128D face embeddings using dlib
- 2Recognition Pipeline: Real-time video capture → face detection (MTCNN) → embedding extraction → nearest-neighbor matching
- 3Anti-Spoofing: Add liveness detection: eye blink detection or depth estimation to prevent photo attacks
- 4Dashboard: Real-time attendance dashboard showing who is present; automatic report generation
Key features to implement
- ✓Real-time face detection in live video feed
- ✓Supports 100+ students with <1s recognition time
- ✓Liveness detection to prevent spoofing
- ✓Automatic attendance report generation
- ✓Admin dashboard with attendance analytics
💡 Unique twist to stand out
Add "Emotion Detection" that logs the general classroom mood over time — showing analytics on average emotion (attentive, confused, tired) during different lecture times.
🎓 What you'll learn
Computer vision, face recognition techniques, real-time video processing, ML model integration, and biometric system design.