PARIS, July 05, 2018 (GLOBE NEWSWIRE) -- The flagship new smartphone OPPO Find X, the latest model of OPPO Find series which had been quiet for four years, was released in the Louvre last week. It not only inherits all the advantages of previous models in the Find series, but also equips with the 3D structured light technology. In collaboration with Megvii, a leading smartphone AI solution provider, OPPO enabled the application of 3D structured light technology in fast unlocking and secure payment, making Find X the world's first Android model to achieve commercial mass production of 3D structured light technology.
Apple released the iPhone X and launched a new feature called "Face ID" for unlocking and Apple Pay based on 3D structured light technology last year. However, all the mass-produced full-screen Android phones launched at that time merely imitated Apple's “bangs” design on the appearance rather than the 3D-structured light-dense camera module behind iPhone X’s “bangs” screen. At a media meeting on May 10 this year, OPPO demonstrated to the public world’s first 5G video call based on 3D structured light technology, and announced that OPPO’s 3D structured light technology was ready for mass production. Today, with the debut of Find X, OPPO not only honored its promise, but also overcome the shortcomings of Android devices in relation to structured lighting technology. The technology is no longer exclusive to iPhone X. Compared to iPhone X, OPPO Find X has abandoned the “bangs” screen design as a result of breakthroughs in hardware and manufacture techniques.
Unlocking + payment, OPPO partners with Megvii to build a model of 3D structured light facial recognition
Compared to traditional facial recognition, OPPO Find X's 3D structured light technology can be applied to a number of scenarios such as secure payment, 3D reconstruction, AR and games. The most important application is secure payment, which makes extremely high technical requirements and needs ultimate quality control on the algorithm used for 3D structured light facial recognition.
OPPO Find X's 3D-structured light facial recognition essentially outperforms all currently marketed models, and it is made possible by Megvii’s 3D structured light facial recognition algorithm.
Here are some features: 0.1s fast recognition, 0.03s reconstruction, 0.0001% false recognition rate, high-precision modeling based on 10000+ security features, and payment-level authentication ...
“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.”
In mid-2017, Megvii began to explore the mobile phone market, and launched a series of mobile AI products such as facial recognition payment and unlocking, portrait lighting effect and background blurring, video editing, 3D Animoji to meet the needs of different mobile phone manufacturers.
In developing the OPPO Find X 3D structured light face solution, Megvii provides three main algorithm model structures for OPPO, including a multi-scale and multi-feature attention detection model, and a life detection model based on ShuffleNet – the mobile convolutional neural network developed by Megvii, and a recognition solution incorporating a self-adaptive network and multiple models. It is reported that in order to meet OPPO's demand, Megvii’s mobile unlocking team worked for several months and completed trainings involving more than 1 billion pieces of facial data and benchmark tests on 600,000 images.
Portrait matting and 3D reconstruction are essential for creating quality portrait light effect
iPhone X portrait lighting technology recently relies on hardware performance and is mainly used in high-end iPhone models with dual camera or structured light, but the final results may not comply with user needs. In particular, the flaws in Apple’s portrait lighting solution become prominent in removing background stage lighting. This is because Apple's portrait lighting effect is processed according to the distance between the target and the camera, so it tends to put things that are on the same level as the portrait in the post-processing of portraits. But this is clearly not what the users want. What they need is enhanced light effect for portraits and facial features.
Unlike Apple, Megvii’s portrait lighting products do not rely heavily on hardware, so they can be widely used in different platforms and models, and can simultaneously meet the requirements of front or rear single/dual camera and can better serve immensely diverse demands on the market.
Technology should be perceived only when it is needed. Therefore, OPPO has always committed to simplicity and using an artistic mind to manufacture premium products. What appear to be simple product designs usually involve more hard work and ingenuity. OPPO Find X is a pioneering Android device with a mass-produced 3D structured light system. As facial unlocking and payment apps become more and more popular, it is foreseeable that 3D structured light facial recognition will soon replace fingerprints, passwords and PIN codes to be the mainstream biometric identification solution for mobile risk management. In the future, the combination of AI and mobile devices will become increasingly close, and OPPO and Megvii will jointly explore the development of technologies and products in the mobile AI industry and achieve deeper market penetration based on secured payment, 3D reconstruction and image processing.
For more information:
Olivia Geng
International Branding Senior Manager
gengfuyao@megvii.com
About Megvii
Megvii is a global leader in computer vision technology. It is committed to empowering different vertical areas with technology by means of “AI+IoT”, providing complete smart solutions for industries. Mobile devices represented by smart phones are the most typical area in which Megvii empowers industry with AI. Therefore, Megvii is not only an industry-based IoT builder driven by AI, but also a core AI solution provider for the mobile phone industry.
