SGL Face Attendance

A local-first attendance workspace that makes the computer-vision decision visible and inspectable.

SGL Face Attendance is a substantially modernized Python project from Self Graphic Lab for person registration, optional face recognition, duplicate-safe attendance events, local SQLite storage, and CSV reporting. It uses synthetic demo data for portfolio review and does not include real biometric samples or cloud credentials.

From camera decision to useful record

The workspace separates configuration, Flask routes, SQLite persistence, optional OpenCV and face-recognition processing, and a responsive operator interface. Frame validation, face-count states, local matching, daily duplicate prevention, search, filtering, and reporting are visible parts of the system rather than hidden inside one desktop script.

Privacy-aware boundaries

The portfolio edition removes the original Firebase persistence layer, committed face images, and service-account expectations. Local records, optional samples, and encoding models stay under ignored application directories. The project is explicit that it has no liveness detection and that confidence is a model output, not accuracy.

View the open-source repository · View SGL DesignScope · Back to Self Graphic Lab