Skip to main content

AI FOR THE FUTURE

  • Live Video Overlays & Match Event Details
  • Automatic Enrollment
  • Events
  • Alarms and Notifications
  • Secure & Private

FACE RECOGNITION AND PERSON-CENTRIC VIDEO ANALYTICS

SAFR from RealNetworks is a visual intelligence platform specializing in face and person-based computer vision optimized for real-world performance.

Building on a 27-year legacy of digital media expertise and innovation, RealNetworks has created a new generation of products that employ best-in-class artificial intelligence and machine learning to enhance and secure our daily lives.

A US-based public company, RealNetworks has maintained a global presence.

“AI FOR THE FUTURE FROM THE PIONEERS OF STREAMING MEDIA”

SAFR for Geutebrück is an AI layer on top of your VMS, providing advanced video analytics that save time, increase efficiency of surveillance operations, and enhance physical security.

SAFR matches faces appearing in live video feeds against watchlist images more effectively, and with less bias, than humans – helping security personnel prioritize feeds that require review and providing them the information they need to respond to persons of interest more quickly.

Transform video feeds into streams of actionable data with SAFR – the best performing face recognition for live video.

Live Video Overlays & Match Event Details

SAFR event details, including live video overlays and side-by-side displays of face recognition events, deliver unmatched efficiency for surveillance operators. Display critical information maintained in SAFR directly on the operator’s screen.

Automatic Enrollment

Easily and efficiently enroll individuals appearing on Geutebrück VMS directly into SAFR’s identity database.

Events

Trigger an event in Geutebrück based on SAFR event details such as name, age, gender, and other parameters. This provides rich, searchable metadata that allows for more efficient post-event investigation.

Alarms and Notifications

Custom alarms instantly notify security personnel when persons of interest approach a monitored area, or when strangers appear in restricted spaces. Easily configure a variety of automated responses to detection and recognition events.

The Best-Performing Face Recognition for Live Video

Transform video feeds into streams of actionable data

With an ever-growing number of surveillance cameras in use, it’s impossible for security professionals to have eyes on every video feed every minute.

Face recognition is a force multiplier.

SAFR matches faces appearing on live video against watchlists more effectively, and with less bias, than humans — freeing up security personnel to manage emerging incidents in real time and giving them the information they need to respond to persons of interest more quickly.

ALWAYS ON, ALWAYS ACCURATE

SAFR uses deep learning to help security teams recognize when an emerging incident would benefit from closer review by a surveillance operator or immediate responsefrom a nearby security officer.

With no gaps in service and a secure central knowledge bank of past incidents and known persons of interest, SARF makes security teams more efficient and effective.SAFR is 99.87% accurate and able to recognize a face on live video in less than 100 milliseconds.

ENHANCED INTELLIGENCE FOR GEUTEBRUCK

SAFR can be deployed as a standalone solution or as asmart layer over GEUTEBRUCK

Reduce alarm fatigue, false positives/negatives, and access-point delays while enhancing situational awareness within a familiar interface.

  • Video Overlays
    Live video overlays automatically identify anyone on custom or third-party watchlists as they appear on video feeds. Track known threats, at-risk individuals, and opted-in VIPs and employees.
  • Configurable Alerts
    Custom alarms instantly notify security personnel when persons of interest enter or leave a monitored area. Easily configure a variety of automated responses to detection and recognition events.
  • Automatic Bookmarks
    Bookmarks containing rich, searchable metadata allow for more-efficient post-event investigation and compliance reporting, and help defend against frivolous lawsuits.

Why SAFR? It works

Not all facial recognition systems are created equal. SAFR can instantly detect and accurately match millionsof faces in real time – even in challenging real-world conditions where faces are in motion, at difficult angles, under poor lighting conditions or partially obscured. NIST tests have consistently found SAFR to be the fastest among the most-accurate systems, as well as among the least-biased algorithms.

This combination of speed, high accuracy and low bias means that SAFR is more likely to return correct matches across faces of all skin tones and genders more quickly than our competitors. Sometimes you get only one look at a face.

With SAFR one look is enough.

Fast & Accurate — SAFR is 99.87% accurate and can detect and match a face in a live video feed in under 100 milliseconds — 3-5 times faster than competing facial recognition algorithms.

Actionable — SAFR provides live security analystics with rich metadata.
View traffic volumes, demographic composition, dwell times and more.
Configure custom reports, alerts and auto-actions based on detection and recognition events — from sending a message to security guards to initiating a building lockdown.

Secure & Private — SAFR uses bank-levelencryption on all face signatures and images in transit or at rest. If run on premises or embedded on device, SAFR passes no data over the internet.
Built with privacy by design principles, SAFR makes it easy to configure data retention settings and manage opt-in/opt-out lists.

Distributed & Flexible — SAFR uses a distributed architecture to consume bandwidth more efficiently than many facial recognition providers.
With the SAFR system’s edge intelligence for detection, recognition on the server, compatibility with off-the-shelf hardware and ability to leverage inexpensive GPUs, adding face recognition to your security tools is easier and more cost effective than you might think.

Deployment Neutral — SAFR can be deployed on a single computer to monitor a handful of video feeds, or scaled to thousands of cameras.
The software can be hosted on premises or in the cloud, or deployed on even the most constrained camera-equipped devices with SAFR SDKs.

Contact us to learn more about our competitively low total cost of ownership.

World-class accuracy and performance where it matters most:
Casinos & Gaming • Airports • Transportation Hubs • Retail • Smart Cities • Law Enforcement