× TECHNOLOGY
SOFTWARE NETWORKING CLOUD IT SERVICES MOBILE SECURITY STORAGE IOT DATA ANALYTICS CYBER SECURITY ORACLE SAP BIG DATA
BUSINESS
DIGITAL MARKETING ERP RETAIL HEALTHCARE TELECOM
MAGAZINE
CURRENT ISSUE ARCHIVE
OPINION ABOUT US CONTACT US CLIENT FEEDBACK

Meet the new trainer for facial recognition

realistic masks train facial recognition

Facial recognition is the in thing right now that’s being used for everything, from unlocking phones, making payment, and even for autonomous cars. But how do you train a device about facial recognition? The answer comes from Japan with a small company called REAL-f Co. making hyper-realistic masks.

Founded by Osamu Kitagawa, the REAL-f Co. was developed with a unique method where high-quality pictures are translated into facial date to make 3D masks. Each mask costs about $2,650 and Kitagawa receives about 100 orders every year. Every mask depicts even the minutest detail of a person’s face like wrinkles, marks, and blood vessels in the eyes!

Real-world IoT applications like using it for medical purposes, developing humanoid robots, improving security measures, etc. are being developed. Japanese car manufacturers are also testing the masks for autonomous vehicles to detect when the driver dozes off!

“I am proud that my product is helping further development of facial recognition technology,” said Kitagawa. “I hope that the developers would enhance face identification accuracy using these realistic masks.” Apple too had developed similar face masks to test its Face ID on the iPhone X. Apple wanted to make sure that false attempts to log into the device can be avoided if the right facial recognition technology could be trained.

MAGAZINES