SPG Attendance : Face Recognition Monitoring

SPG Attendance Monitoring uses face recognition to track field staff attendance across multiple store outlets. It combines face matching with location data and provides a real-time dashboard plus Telegram notifications for supervisors.

Manual attendance tracking for distributed field staff is unreliable. Phone-based check-ins can be faked, and supervisors can't verify attendance at multiple outlets simultaneously.

Build a reliable attendance system that uses face biometrics and location context so supervisors can trust the data without being on-site.

• Face detection and encoding during staff enrollment.
• Attendance check-in via mobile with face capture.
• Backend matching service compares captured face against enrolled database.
• Dashboard shows real-time attendance per outlet with status flags.

• Staff enrollment with face registration.
• Face matching for daily check-in with confidence scoring.
• Location metadata alongside face verification.
• Telegram notifications for supervisor alerts.

• Face detection and alignment from phone-captured images.
• Face encoding and similarity matching.
• Confidence thresholding configurable per deployment.

1. 1Staff enrolls face through the registration flow.
2. 2Daily check-in: staff captures selfie at outlet location.
3. 3API runs face detection → encoding → matching against enrolled set.
4. 4Match result + location metadata stored as attendance record.
5. 5Supervisor dashboard displays real-time status per outlet.

• Varying lighting conditions and phone camera quality across outlets.
• Handling false rejections when staff appearance changes.
• Ensuring the face capture happens at the actual outlet location.

• Combine face matching with location data rather than relying on biometrics alone.
• Use notification-based exception reporting instead of real-time monitoring for every check-in.
• Keep manual override path for legitimate edge cases.

Working prototype with face enrollment, daily check-in matching, and supervisor dashboard. Deployed for pilot testing.

• Add liveness detection to prevent photo spoofing.
• Add offline check-in mode for areas with poor connectivity.
• Build per-outlet attendance reports.