Use Cases
End products you can build with embedded biometrics. Each use case implements the Identification (1:N) / Verification (1:1) and Enrollment features and maps them to concrete SDKs; pick by where recognition runs and the form factor.
- On-device identification — the whole recognition pipeline runs locally on the device; offline, and biometrics never leave the device. Face, palm, fingerprint.
- Server-side identification — a client sends a probe to a central service that matches against a server-side watchlist; centrally managed, large galleries. Face, palm.
On-device vs server
The same recognition can run on the device or on a central server:
| On-device | Server-side | |
|---|---|---|
| Where recognition runs | On the device | Central server |
| Connectivity | Works offline | Requires a network path |
| Biometric data | Stays on the device | Sent to the server |
| Gallery | On-device (bounded) | Server-side watchlist (large, API-managed) |
| Modalities | Face, palm, fingerprint | Face, palm |
Prefer on-device for offline operation, data locality, or constrained devices; prefer server-side for large or centrally-managed galleries and many thin clients.