Face Recognition - Auto-Enrollment

Streamline your access control and visitor management with Scylla’s Auto-Enrollment, a fully automated solution driven by AI.

Book a demo

Related materials

How to Minimize the Hidden Costs of Physical Security Breaches with AI

Facial Recognition: Practical Applications for Physical Security

How Facial Recognition Transforms Access Control in the Middle East

FAQ

The face recognition auto-enrollment system operates on two databases. The first one's a vector database that stores generated face vectors, which are non-reversible, meaning the source image cannot be reconstructed from them. The second one's a conventional database that stores the actual images. Storing images is optional, and the facial recognition system can function solely with the vector database if preferred by the client.

Yes, a pilot program can be conducted. Success is measured by the value of statistical insights derived using the BI dashboard and forensic search features. Note that the accuracy of the system may be lower due to its fully automatic nature compared to the watchlist system that we offer.

For Full HD streams, auto-enrollment requires 500 PPM (pixels per meter), which typically corresponds to a detection range of approximately 2 meters (with closer distances yielding better results). For subsequent recognition after enrollment, 250 PPM is sufficient, typically corresponding to a range of about 4 meters. Latency ranges from 3 to 4 seconds, depending on the specific system setup.

The number is configurable and depends on the system's hardware. Increasing the number of faces to be analyzed requires more robust hardware.

Yes, Scylla face recognition system performs efficiently even with significant transformations or changes in appearance. However, for re-identification, the person must have been successfully enrolled in the database. During enrollment, stricter criteria are applied, meaning partially obscured faces are skipped to ensure high-quality captures in the database.

Scylla face recognition watchlist focuses on accurate identification of pre-enrolled individuals for immediate action. The enrollment is manual and curated for real-time recognition of flagged individuals. It's an ideal solution for airports and other transportation hubs, law enforcement and event security. Scylla auto-enrollment solution excels in high-traffic environments such as casinos and retail, by dynamically tracking individuals and creating actionable data for behavioral insights. The database is built automatically without the need for any manual input. It's a perfect fit for casinos, retailers and corporate campuses.