EduTrack - Image-Based Recognition Attendance System
Ashish Kumar
Published
0
• Built a full-stack attendance system connecting MTCNN/FaceNet inference pipelines to the backend via REST, achieving 92% face-recognition accuracy with sub-500 ms response times under live load.
• Designed a three-tier RBAC system (Admin, Faculty, Student) with fine-grained permission logic, cutting manual attendance recording time for a 60-student cohort by approximately 80%.
• Secured the application with stateless JWT authentication and bcrypt password hashing; wrote automated Postman test suites with zero unauthorized-access incidents during testing.