“SkyBiometry is pitching the API for user authentication, access management, and attendance tracking, as well as interactive marketing and database management for companies who need to sort through a large number of photos…”
The Neurotechnology subsidiary SkyBiometry has improved its facial recognition API with an upgraded algorithm that allows the platform to detect and recognize faces faster and more accurately. The new algorithm has performed well on the NIST’s Face Recognition Vendor Test, achieving a top-15 score in the ongoing “wild” test and a top-6 score in the largest test in 2018.
According to SkyBiometry, the new version of the algorithm is able to read faces at challenging angles (including a full profile), and categorize people based on a wider array of facial features. Previous versions of the algorithm could pick out certain emotions and attributes like age and gender; now, the latest version can handle hats and facial hair, with SkyBiometry boasting that it can also determine someone’s ethnicity.
SkyBiometry is pitching the API for user authentication, access management, and attendance tracking, as well as interactive marketing and database management for companies who need to sort through a large number of photos.
Neurotechnology is best known for the multimodal MegaMatcher solution, and released a new facial recognition SDK at the beginning of the year. The company’s attribute recognition technology is also featured in SentiMask 2.0.
–
July 16, 2019 – by Eric Weiss
Follow Us