Cyberlink scales up video monitoring with upgraded facial recognition – Technologist

Taiwan-based CyberLink, a pioneer in AI and facial recognition technologies, has rolled out a major upgrade of its FaceMe Security solution, adding features such as people identification and contactless access control.

FaceMe is optimised for IoT devices, being uniquely positioned to integrate edge-based AI facial recognition into a wide range of IoT and AIoT solutions. FaceMe is optimised to run on most hardware configurations, from high-end workstations to low-power chipsets typically used in IoT devices. It is compatible with Windows, Linux, Android and iOS systems.

The solution latest updates include the ability to identify people with a high degree of precision even when wearing a mask, as well as mask detection and temperature measurement for health control. The solution now offers enhanced compatibility to a number of mainstream video management systems (VMS) and further optimisation to chipset support, including the cost-effective NVIDIA Jetson platform.

“Biometric and vision technologies are setting new standards for IP surveillance deployments. Not only can solutions using top-tier facial recognition accurately verify identity, even for faces partially covered by a mask, but they can also detect if the mask is worn properly,” said Dr. Jau Huang, CEO of CyberLink.

The solution comes with all the features needed to enable automated and contactless security monitoring, access control and health checks, each of which are relevant to organisations of all types and sizes, in all sectors of the economy.

Small business owners can literally setup affordable single-computer, single-camera systems on their own. The solution’s robust and scalable architecture enables a rapid and easy deployment by system integrators anywhere, from single-location businesses to very large organisations, addressing use cases relevant across a wide range of industries and sectors, including retail, manufacturing, warehousing, office management, hospitality and many others.

Its flexibility, such as the ability to connect into existing IP cameras and run on multiple types of computers and workstations, often allows integrators to implement FaceMe Security across entire organisations, all from the server room. The versatility of its features and a collection of APIs ensure a seamless integration into systems such as visitor management, employee time and attendance, access control and automated door operation, in addition to connecting easily into existing surveillance and monitoring systems.

All-inclusive solution

The newest edition of FaceMe Security Workstation allows for real-time face detection and facial template extraction at the edge. Optimised to run across a wide range of edge-based hardware, FaceMe Security Workstation can be deployed on high-end Windows workstations equipped with single or multiple NVIDIA Quadro GPUs for heavily trafficked areas of up to 80,000 people per hour. For mid-sized sites with traffic ranging from 2,000 – 6,000 people per hour, such as a warehouse, FaceMe Security Workstation can be deployed on NVIDIA Jetson AGX Xavier or Xavier NX, providing modest levels of performance with a significantly lower financial and energy costs.

“FaceMe Security is an all-inclusive solution that can completely overhaul existing IP surveillance infrastructures to deliver the latest security, access control and health screening capabilities, for organisations of all sizes, across all industries,” Huang said.

Furthermore, FaceMe Security’s notification API supports mainstream VMS (video management systems), including Milestone, Network Optix Nx Witness and VIVOTEK VAST2. It enables real-time detection of VIP, block-listed and tagged personnel entering surveillance zones and triggers a notification to the VMS, informing security personnel of real-time events.

In addition to its built-in features, FaceMe Security offers two add-ons that expand its capabilities to provide complete time and attendance as well as integrated health screening. The FaceMe Security Check-In add-on can be deployed on light-weight PCs at building entrances for real-time, on-screen identity verification along with check-in information. The FaceMe Security Health add-on provides an inconspicuous health-monitoring solution for any facility. The AI engine detects if a person is wearing a mask properly over the nose and mouth, verifies their identity, and measures their body temperature. When detecting a person with elevated body temperature or someone not wearing a mask, a notification including the person’s picture and location is sent instantly to security personnel for further actions.