3rd YearAdvancedAI/ML 6-8 weeks
Face-Recognition Attendance System
An attendance system that detects and recognises faces from a webcam, marks attendance automatically, and exports daily reports.
PythonOpenCVface_recognition / FaceNetFlaskSQLite
Find GitHub repos & source code →How to build it — step by step
- 1Enrolment: Capture face images per person and compute and store face embeddings.
- 2Recognition: Detect faces in the video stream and match embeddings within a distance threshold.
- 3Attendance logic: Mark present once per session, preventing duplicates; flag unknown faces.
- 4Reporting: Persist records and export attendance as CSV/PDF per date and class.
Key features to implement
- ✓Webcam face enrolment
- ✓Real-time recognition
- ✓Duplicate-proof marking
- ✓Unknown-face alerts
- ✓CSV/PDF reports
💡 Unique twist to stand out
Add a basic liveness/anti-spoofing check (blink or texture analysis) so a printed photo cannot mark attendance.
🎓 What you'll learn
Computer vision, face embeddings, similarity thresholds, and integrating ML into a web app.