FB pixel

Megvii algorithms power 3D facial recognition in new OPPO smartphone


Megvii is providing facial recognition for fast unlocking and secure payments with the new Find X smartphone from OPPO, which says it is the first Android model commercially mass produced with 3D structured light technology.

OPPO demonstrated a 5G video call based on 3D structured light technology in May, and its technology can be used for 3D reconstruction, AR and games, in addition to unlocking and payments. The new Find X’s 3D camera, which is not housed in a “notch,” provides recognition in 0.1 seconds, reconstruction in 0.03 seconds, a false positive rate of 0.0001 percent, high-precision modeling based on more than 10,000 security features, and payment-level authentication, according to the announcement.

“Sixty percent of the world’s cameras are used on mobile phones,” said Wu Wenhao, Senior Vice President of the Cloud Computing Technology Department of Megvii, “In a short term, the priority of Megvii is to empower the world’s one billion cameras. So the mobile phone market is meant to be the main theater of war.”

Megvii is providing OPPO with three main algorithm structures, including a multi-feature and multi-scale attention detection model, a liveness detection model based on its ShuffleNet mobile convolutional neural network, and a recognition system with a self-adaptive network and multiple models. The company says that its portrait lighting products differ from those from Apple in that they do not rely heavily on hardware, allowing them to be used in different platforms and models to serve diverse market demands.

MIT Technology Review recently named Megvii founder Yin Qi to its Innovators Under 35 2018 list, and the company began a push into the Southeast Asian market with a distribution deal for Thailand.

Related Posts

Article Topics

 |   |   |   | 

Latest Biometrics News


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Most Read This Week

Featured Company

Biometrics Research

Biometrics White Papers

Biometrics Events

Explaining Biometrics