Face Recognition

Every Face Has a Story

Face Recognition is a Senstar Symphony-based video analytic that identifies known and unknown individuals. Using a combination of patented 2D to 3D pose correction technology, this analytic is designed for fast, reliable identification under real-world challenges, including lighting, angles, facial hair, pose, glasses and other occlusions, motion, crowds, and expression.    
Thumbnail Video Image

Fast Image Processing

Senstar Face Recognition has registered a template correlation speed of 25 million matches per second (.040 microseconds/match), per core, against both live and archival video. This makes facial biometric, multi-factor authentication possible and translates to accelerated response times for security and customer relationship applications.

Accurate, Reliable Detection

Face Recognition delivers results in less than ideal conditions. It performs in crowds, where faces are partially obstructed, when facial appearance or expression has changed, and includes liveness checks to prevent spoofing. It has achieved a 99.44% Rank 1 matching accuracy on the Facial Recognition Technology (FERET) high pose data set and a 97.60% Rank 1 matching accuracy on the Labeled Faces in the Wild (LFW) data set, with a false positive rate of 0.1%. You can rely on Face Recognition for the most demanding applications.

Smart Search

Perform forensic investigations by quickly searching historical video archives for known and unknown people. Symphony Face Recognition search works across multiple cameras, with search results being filterable by match score or date/time. Video results can easily be exported for use by authorities or management.

Improve Building Security

Built-in Symphony, Senstar Face Recognition results can be integrated with Symphony and third-party access control systems, enabling organizations to implement two-factor authentication and automated access processes.

Identify Known Offenders

Improve security in retail, border security, and educational environments by generating alarms whenever known offenders are present. Alarms can be sent via email or SMS messages, ensuring on-site staff are aware of the offender’s presence.