A camera-based attendance system that automatically records student and professor presence using face recognition — reducing paperwork and manual effort for every session.
From the moment the camera goes live, the system handles detection, logging, and status marking — so you don't have to.
Uses InsightFace's buffalo_sc model to identify students and professors from live CCTV footage — fast and angle-resistant.
Attendance records are written to the database the moment a face is confirmed — with timestamps accurate to the second.
Automatically marks students as PRESENT or LATE based on the professor's session start time and a configurable grace period.
Switch seamlessly to event attendance tracking for school-wide gatherings — automatically detected from the schedule database.
Sessions are anchored to the professor's recognized face. Students are only recorded while the professor's session is active.
Works with IP cameras over RTSP — including Tapo and other network cameras — with automatic reconnection on stream loss.
The camera detects and recognizes the professor's face, automatically creating a session log and activating student recording.
As students enter the room, their faces are matched against the database and attendance is instantly written with PRESENT or LATE status.
When power cuts or a manual stop is triggered, the session is closed gracefully and all records are preserved in the database.