SAFR For Security Provides Accurate Real-Time Face Recognition

Download PDF version

Facial recognition has seen huge breakthroughs since the U.S. National Institute of Standards and Technology (NIST) first began testing in 2010. Accuracy has seen massive gains, especially from 2013-2018.

In the 2018 test, the most accurate algorithm was 20 times more accurate than the 2013 equivalent. Essentially, 95 percent of the matches that failed in 2013 now yield correct results. Compare that to 2010-2013, when the most accurate algorithm reduced its error rate by 30 percent. This reduction in error rates since 2013 is due to wholesale replacement of the old algorithms with new ones based on deep convolutional neural networks — completely revolutionizing the technology.

Optimal recognition results

SAFR says it delivers optimal recognition results with 99.86 percent accuracy in under 100 milliseconds

One entrant in the newly energized market is RealNetworks, whose SAFR for Security is an AI-based facial recognition solution for live video that integrates video management system (VMS) solutions. With 24/7 monitoring, SAFR detects and matches millions of faces accurately in real time, enabling teams to manage a watchlist across any number of video feeds.

SAFR says it delivers optimal recognition results with 99.86 percent accuracy in under 100 milliseconds, even in real-world conditions where faces are in motion, at different angles, under poor lighting, or partially obscured. SAFR builds on RealNetworks’ 23-year history in video technologies. Launched in July 2018, SAFR — secure, accurate facial recognition — is enabling new applications for security, convenience, and analytics.

Create security responses

“We seek to be the world’s most trusted facial recognition platform and are delighted to partner with customers in the security industry and elsewhere to shape a more secure, convenient future worldwide,” says Dan Grimm, Vice President of Computer Vision and General Manager of SAFR at RealNetworks. “Security professionals are asked to keep us safe 24/7, monitoring a burgeoning number of cameras, and we help make them more effective.”

SAFR targets facial recognition for live video, identifying camera-unaware faces moving in real-world conditions. In the April 2019 NIST results, SAFR tested as the fastest and most compact solution among algorithms with less than 0.022 False Non-Match Rate — 62 percent faster than the average speed, according to the company. SAFR now provides capabilities such as live video overlays alerting security professionals to events in real time, automatic bookmarks with rich metadata for investigative work, and alerts that can be customized to create security responses.

SAFR says they have achieved a balance of accuracy and performance for live video

SAFR uses one-sixth the compute power of competing facial recognition solutions

Facial recognition algorithms

Five years ago, facial recognition algorithms would struggle to match forward-facing people from still images, let alone camera-unaware moving faces from live video with variations in rotation and tilt. SAFR says they have achieved a balance of accuracy and performance for live video. A contributor to this accuracy is consistency across a range of skin tones. The algorithm was trained on a highly diverse global set of over 10 million non-simulated real-world faces.

SAFR was optimized for speed and can sample a face multiple times during the same period of time as other algorithms, subsequently increasing its accuracy. SAFR achieves the performance through edge processing. Distributed architecture enables efficient bandwidth consumption, reducing the roundtrip latency of facial recognition speed to under 100 milliseconds. The savings lower total cost of ownership (TCO): SAFR uses one-sixth the compute power of competing facial recognition solutions, equating to $500,000 or so in savings on a 250-camera deployment.

Integrated experience

SAFR also uses off-the-shelf hardware and is optimized to leverage inexpensive GPUs

SAFR also uses off-the-shelf hardware and is optimized to leverage inexpensive GPUs. SAFR can be deployed on premises or in the cloud, and supports Windows, Linux, macOS, iOS, and Android. When SAFR is paired with a VMS, such as Milestone XProtect or Genetec Security Center, the integrated experience includes 24/7 monitoring to detect and match faces in real-time.

Features include live video overlays within the VMS to identify strangers, threats, concerns, unrecognized persons, VIPs, employees, or other tagged individuals in live video. Real-time alerts can be customized for when persons of interest appear on a video camera feed. Additionally, automatic bookmarks with rich metadata make for easier investigative review of security footage. Facial recognition technology is increasingly in demand to improve safety across various industry verticals.

Better customer experience

Large enterprises with high-visitor flows and heightened security — such as transportation hubs, stadiums, universities, and hospitals — need to know in real time when persons of interest or those on watchlists appear on camera. Sports stadiums could apply facial recognition to deny entry to banned patrons, locate lost children, or recognize VIPs to deliver a better customer experience.

Hospitals need access control to restricted areas and pharmaceutical storage closets

Hospitals need access control to restricted areas and pharmaceutical storage closets. Airports and transit centers value traffic flows, demographic composition, and dwell times to help improve scheduling. SAFR for Security is available worldwide, and the company partners with VMS providers such as Milestone, Genetec, Digifort, and IPConfigure by Paliton Networks. They are actively working to support additional VMS solutions and have sales teams located in major metropolitan cities around the world.

Security professionals

“The job of the security professional is critical in today’s world,” says Grimm. “SAFR for Security helps mitigate the challenges of the important work security professionals do to keep us all safe.” In designing and developing SAFR, RealNetworks considered diversity and the uniqueness of each person; Grimm says their massive global training data set is a competitive advantage.

SAFR is designed with privacy in mind. All facial images and signatures are AES-256 encrypted in transit or at rest. “SAFR is powerful enterprise-grade software that is continuously improving through innovation and many years of expertise,” says Grimm.

Download PDF version

Author profile

Larry Anderson Editor, SecurityInformed.com & SourceSecurity.com

An experienced journalist and long-time presence in the US security industry, Larry is SecurityInformed.com's eyes and ears in the fast-changing security marketplace, attending industry and corporate events, interviewing security leaders and contributing original editorial content to the site. He leads SecurityInformed's team of dedicated editorial and content professionals, guiding the "editorial roadmap" to ensure the site provides the most relevant content for security professionals.

People mentioned in this article

Dan Grimm RealNetworks Inc.

In case you missed it

How Does Security Technology Make Our Schools Safer?

The new school year is a good time to reflect on the role of security in protecting our schools. From video to access control to some newer technologies, our Expert Panel Roundtable found plenty to talk about when we asked this week’s question: How does security technology make our schools safer?

Security Device Manufacturers Develop Interfaces To Be Compatible With Gunshot Detection Technologies

Gunshot detection today is part of more physical security systems than ever before, and many manufacturers are developing interfaces to the latest gunshot detection technologies. Genetec has integrated ShotSpotter gunshot detection technology into its unified IP security platform, Security Center. Thanks to this integration, police departments and security professionals will be able to receive more actionable information, gain rapid access and detailed location insights when a gunshot situation occurs. Cloud-Based analysis software ShotSpotter uses wide-area acoustic surveillance that automatically gets activated when gunfire occurs ShotSpotter uses wide-area acoustic surveillance that automatically gets activated when gunfire occurs, providing comprehensive outdoor coverage on campuses and in complex urban geographies. ShotSpotter sensors and cloud-based analysis software triangulate and pinpoint the precise location of each round fired within seconds, the number of shots fired and provide an immediate gunfire alert with an audio clip. Data capture form to appear here! Detailed incident data is immediately sent to SST’s Incident Review Center (IRC), a secure data processing and alert validation facility, where acoustics experts analyze the data and qualify the incident in real-time 24/7. Once validated, an alert is sent directly into Genetec’s Security Center, all in under a minute. Once the alarm is triggered within Security Center upon gunshot detection, operators can quickly access video and other related data surrounding the location of a gunshot alert, and immediately visualize it on Security Center maps, leveraging the dynamic visualization capabilities of the Security Center Plan Manager. Genetec has integrated ShotSpotter gunshot detection technology into its unified IP security platform Gunshot identification software Stanley Security has announced a partnership with Shooter Detection Systems LLC (SDS), gunshot detection solutions provider. As an authorized dealer, STANLEY Security is now certified to sell, install and service SDS products and services. SDS’s Guardian Indoor Active Shooter Detection System combines acoustic gunshot identification software with infrared gunfire flash detection for a fully automated gunshot detection and alerting solution. The Guardian System immediately detects gunshots and simultaneously alerts building occupants The Guardian System immediately detects gunshots and simultaneously alerts building occupants and first responders within one second and with zero false alerts. Johnson Controls has announced that its exacqVision video management system now integrates with Databuoy Corp.’s ShotPoint acoustic gunshot detection and localization system to speed response time and heighten the decision-making process for local responders during active shooter situations. Integrated security systems Johnson Controls has also released the Detect360 Active Shooter Response (ASR) system, combining reliable gunshot detection with notification technology to provide immediate warning when a gunshot is fired within a building. By providing early notification and precise location of the shooter, the ASR system gives occupants time to find safety and allows police and security personnel to mitigate the threat up to 60 percent faster. Detect360 ASR gives first responders information as it develops at the scene, including shot location, number of shots, location of the shooter in near real-time and the ability to remotely monitor the situation through integrated security systems. It also collects forensic information that can later be used by law enforcement. The system supports multiple integration options, including video feeds from existing CCTV and intrusion systems, mass notification systems, access control systems and panic button/manual initiation systems. Detect360 ASR gives first responders information as it develops at the scene Public safety access points Listed as one of the most important new technologies debuted at CES 2019 by the Wall Street Journal and NBC News, Safe Zone also showcased its Safe Zone Gunfire Detector (GFD) at ISC West 2019. Priced at $149 per unit, the Safe Zone Gunfire Detector combines infrared and sound detection technologies with immediate cloud-based data analysis that can analyze more than 3,000 data points of each loud noise that exceeds ambient levels by a certain threshold. The system determines the location of shots fired, the number of shots, and the type Utilizing the data from multiple detectors simultaneously, the system determines the location of shots fired, the number of shots, and the type and caliber of gun being used. Within 10 seconds of the trigger pull, an alert is sent to Public Safety Access Points in the appropriate dispatch center, giving local 911 dispatchers all critical information on their computers. Reseller recruitment program Shot Tracer has launched a reseller recruitment program with a highly attractive on-boarding proposition for qualified dealers. Qualifying resellers can receive a Gunshot Simulator Kit valued at $1,000 when they commit to becoming a Shot Tracer dealer. Shot Tracer’s Eagle Gunshot Detection System easily integrates with virtually any security, alarm, surveillance and access system via contact closure – or wirelessly via IP integration. In fact, Shot Tracer gunshot detection sensors are as easy to install as a smoke detector, so there’s virtually no learning curve for installing dealers. The new Shot Tracer Reseller Recruitment Program includes an exclusive offer for installing dealers to receive a free Gunshot Simulator Kit, valued at over $1000, which is essential for demonstrating Shot Tracer Eagle Gunshot Detection sensors. Terms and conditions apply for qualifying resellers.

viisights’ Next Generation Of Video Analytics Offer ‘Behavioral Understanding’

A video analytics system that provides ‘behavioral understanding’ can yield more meaningful and actionable data for a range of applications. In public safety and security, such a system can alert on violent or suspicious behaviors, such as people fighting, vandalism, people with weapons, etc. In advanced traffic surveillance and monitoring, it can provide alerts to vehicle collisions (accidents), traffic hazards or vehicle that aren’t using the road properly, such as a car that stops in the middle of the junction. For enterprise and campus security, it can provide advanced anti-tailgating and detect unauthorized activity. Video surveillance infrastructure viisights was founded by a group of entrepreneurs with track records in developing technology businesses These uses are among the benefits of viisights’ video analytics technology based on behavioral understanding of video content. “It means we can extract more meaningful data from the huge amount of video content that is captured, and we can transform that data to actionable insights that eventually justify the massive investment in video surveillance infrastructure,” says Asaf Birenzvieg, CEO of viisights. Their behavioral understanding systems for real-time video intelligence leverage artificial intelligence technology. viisights was founded by a group of serial entrepreneurs with track records in developing technology businesses. The Israeli company’s founders recognized a growing global need for intelligence to make physical and virtual public areas safer – and realized the role that smart video understanding technology can play. Developing artificial intelligence technologies viisights is committed to developing artificial intelligence technologies that facilitate human-like video understanding, which in turn serves as the basis for fully autonomous video intelligence systems powered by pattern prediction technology. “Behavioral recognition is the future of video analytics and the next generation of the object classification analytics systems that hold the majority of the market today,” says Birenzvieg. viisights has developed a video understanding technology for real-time video processing “To date most video analytics systems still base their product features on static analysis of objects from images using image recognition, even the ones that use ‘AI analytics.’ Products built using such object classification technology are extremely limited.” For example, object classification analytics cannot recognize behavioral events in a video such as people fighting or a car collision because such behaviors can’t accurately be concluded in large scale from analyzing a single static image/frame. Video understanding technology viisights has developed a video understanding technology for real-time video processing. The technology can process live video feeds. In addition to recognizing a particular object (e.g., person) and its attributes (e.g., red shirt), the system can understand an object’s actions, interactions with other objects (events), the scene being viewed (i.e., crowd is gathering, riots) and the context (a car is driving on the road or on the sidewalk). The main verticals are smart cities, enterprises and campuses, banks and ATM security“Basically, we are able to extract more meaningful data from a live video feed and therefore create actionable insights and greater ROI,” says Birenzvieg. The company focuses mostly on security and safety use-cases. The main verticals are smart cities, enterprises and campuses, banks and ATM security, security guard companies and transportation hubs. The company is working on a new product for in-vehicle monitoring mostly for security, safety, vehicle protection and proper vehicle use; it monitors passengers’ behavior inside a bus, train, or taxi. The product will come to market next year. Video management system viisights’ video analytics offering is currently optimized for server-side deployment, and the integration architecture is similar to most video analytics systems. From one side it is integrated with the video management system (VMS). They are a Milestone verified partner and soon will be part of Milestone's marketplace. From the other end, it is connected to a command-and-control system for processing the data and presenting the alerts to the end-user. The analytics company makes most sales through system integrators. They have partnerships with big system integrators like Motorola Solutions and NEC and are also working with smaller ones. They are looking to expand their system integrator network, mostly in the USA and Europe. Behaviors can have many variations and they can be very diverse Cloud video surveillance “We will continue to invest in performance and accuracy, meaning higher recall and lower false positive rate,” says Birenzvieg. “Since our major value proposition is in behavior recognition, behavior events many times are not clearly defined, which is very different from object classification. Behaviors can have many variations and they can be very diverse.” An example is a simple behavior like a person falling on the floor. A person can fall on the floor in many ways, but the challenge is to ignore similar behaviors that are not a person falling and that confuse the system, such as a person bending over to tie his shoelaces. With cloud video surveillance becoming a trend, viisights is also looking into offering some of their advanced functionalities in a video-analytics-as-a-service-model.