CyberLink will be demonstrating AIoT solutions powered by FaceMe, its new facial recognition engine, at CES 2019. FaceMe provides Smart Retail solutions to analyze store traffic and visitors’ attributions, Smart Security systems with anti-spoofing capabilities, and Smart Office solutions, such as background blur that allows business professionals increased privacy during video conferences.
Endorsed by the world-renowned MegaFace Challenge, FaceMe is one of the world’s top facial recognition engines, powered by deep learning and neural network algorithms. Its high precision yields up to a 98.41% true acceptance rate (TAR) with a false acceptance rate (FAR) of less than 10-6 (0.000001). FaceMe provides excellent cross-platform capabilities that support Windows, Linux, Android and iOS for a variety of hardware configurations, such as digital signage, CRM solutions, Kiosks, access and security systems, and personalized robots.
“With more than five years of research and development in AI facial technology, FaceMe is ranked as the top-class facial recognition solution for multiple AIoT applications,” said Dr. Jau Huang, CEO of CyberLink. “To achieve advanced cross-platform integration, FaceMe is also optimized for edge computing devices, which help hardware providers implement FaceMe into existing systems for lightning-fast response times and enhanced performance.”
At CES 2019, CyberLink will demonstrate the uses of FaceMe in Smart Retail, Smart Security and Smart Office scenarios. For Smart Retail, by implementing FaceMe on digital signage and KIOSKS, stores can track customer traffic and analyze demographics such as gender, age and emotion. The analysis helps enhance the shopping experience and customize promotions in real time.
For Smart Security solutions such as door access controlled by embedded systems, or surveillance cameras connected to VMS (Video Management System), FaceMe helps hardware developers deploy advanced facial recognition and identification capabilities. FaceMe is also able to identify certain facial expressions and detect potential identity fraud with depth information produced from a 3D camera.
