Personal / study project
Face Recognition Attendance Demo
A personal study project exploring real-time face detection and recognition for lightweight, camera-based attendance tracking.
Attendance log (demo data)
Illustrative interface with fictitious sample data, shown in place of confidential client screens.
Why I built it
A hands-on way to get comfortable with classical computer-vision pipelines alongside my production data-engineering and backend work: face detection, embedding-based recognition, and serving predictions behind a simple API.
How it works
OpenCV handles frame capture and face detection; the face_recognition library (built on dlib) generates face embeddings and matches them against a small enrolled set. A thin FastAPI service exposes an endpoint that accepts a frame and returns matches with a confidence score, which a small front end renders as bounding boxes and an attendance log.
Status
A working local demo rather than a production system. The screen above shows the UI with entirely fictitious names and timestamps.