With SAFR SC800 Facial Recognition Camera, SAFR engine is embedded directly into the camera and utilizes onboard compute functionality — eliminating the need for servers dedicated to processing video streams. By eliminating server hardware, SAFR SC800 saves end-users on average $500 per camera in on-premise server hardware.

SAFR SC800 Facial Recognition Camera runs SAFR Face Recognition algorithm on the edge. The SAFR algorithm provides the following performance:

- Accuracy: 99.87% precision in LFW faces FNMR

- Accuracy with masked faces: 96.16% TPR at 0.0047% FPR on masked faces

- Face match for Live Watchlist: < 100ms Average Response Time

- Face search / Forensics analysis: 500,000 events/second

SAFR SC800 Facial Recognition Camera embeds the complete face recognition engine with a watchlist of up to 50k people of interest, without the need of external processing servers. SAFR SC800 performs the entire facial recognition process on the edge, including:

- 4k video processing

- image optimization (face prioritization auto-exposure)

- face capture - biometric signature

- database match

In case of unstable network connections, SAFR SC800 can operate offline with the ability to store up to 20,000 events offline.