logo
Image comparison
Administrator login
Menu

Inspiration

IDent began as a vision to make identity verification faster and safer. It evolved from facial recognition into a complete access-control platform combining live detection, image search, enriched profiles, analytics and audit history.

What it does

IDent identifies people in real time through facial recognition or uploaded images. It compares faces with registered profiles, shows access status and confidence scores, and helps administrators manage identities, alerts, documents and detection history.

How we built it

We built IDent as a modular desktop application with Python and CustomTkinter. OpenCV handles camera capture, face detection and encoding; a matching engine compares faces with enrolled profiles; MySQL stores identities, permissions and detection logs; and a responsive UI connects authentication, dashboards, profiles, image search and analytics.

Challenges we ran into

We faced challenges with real-time camera performance, face-detection reliability, database connectivity, and keeping recognition results stable. We also had to manage sensitive biometric data, prevent false matches, integrate multiple UI workflows, and maintain a consistent experience across different screen sizes and hardware conditions.

Accomplishments that we're proud of

We built a complete identity platform instead of a simple facial-recognition demo. IDent supports real-time identification, multi-pose enrollment, imported-image comparison, enriched profiles, access statuses, analytics and detection history. We are especially proud of creating a clear end-to-end workflow that connects recognition technology with practical security operations.

What we learned

We learned that building reliable facial recognition requires more than accurate matching. It also depends on image quality, stable camera processing, clear confidence thresholds, strong database design and careful handling of biometric data. We also learned the importance of modular architecture, user-centered workflows, responsive interfaces and human review for uncertain results.

What's next for IDent

Our next steps are to strengthen liveness detection, improve recognition accuracy, add stronger privacy controls and complete the imported-image comparison workflow. We also plan to support multiple cameras, real-time alerts, better reporting, a web dashboard and more scalable face-recognition models for production use.