How does face verification stop proxy candidates at exam centres?
Face verification compares the live face of a candidate at the exam centre entrance with the photo registered for that candidate. With passive liveness, it also rejects photos and screen replays. Done on a phone or tablet, it works offline, needs no special hardware and can run across hundreds of centres at once.
Approaches compared
| Criteria | Manual ID check | Webcam-based online proctoring | Dedicated camera gates | Mobile-first edge verification (FaceTagr) |
|---|---|---|---|---|
| Works with poor connectivity | Yes, but slow and inconsistent | Typically needs internet | Depends on setup | Yes, runs on the device |
| Hardware spend | None | None | Typically significant | None, uses existing phones and tablets |
| Scales across many centres | Hard | For remote candidates | Costly per centre | Distribute the app |
| Handles old or low-resolution ID photos | Relies on human judgement | Varies | Varies | Designed for it, proven in Bihar elections |
| Audit trail | Paper | Yes | Yes | Yes |
What to look for
Frequently Asked Questions
Does it work offline?
Yes. Face matching and passive liveness run on the device. Records queue and sync when connectivity returns.
Does it work with low-resolution or old ID photos?
Yes. This is a core strength, proven in the Bihar elections against decades-old voter photographs.
What devices do we need?
Android or iOS phones or tablets. Android needs 2GB RAM minimum. No special cameras or servers.
Can it connect to our registration or exam system?
Yes, through a RESTful API, and through an SDK for native, Flutter and React Native apps.
How is this different from online proctoring?
Online proctoring watches remote candidates through a webcam. FaceTagr verifies identity at the point of entry at physical exam centres, and the SDK can be embedded in your own candidate app. It complements online proctoring.
Can we start with a pilot?
Yes. Start with a single exam centre and scale from there.