3DiVi: The Next Generation Expert in Face and Gesture Recognition

Dmitry MorozovDeep learning is an aspect of artificial intelligence (AI) that emulates the human learning approach that is used by the mankind to gain knowledge. Put into simple words, deep learning is deployed to automate predictive analytics. To understand the complexities of deep learning, imagine an automated vehicle which learns to identify routes by constant monitoring at the maps. It learns and relearns the routes and traffic conditions. What an autonomous vehicle does here without knowing is clarifying a complex abstraction (the concept of correct routes) by building a hierarchy of each level of abstraction with the knowledge that was learned and re-learned from the preceding layer of the hierarchy.

Deep learning is deployed by computer algorithms that go the same process what the automated vehicle does to accurately predict outcomes. Each algorithm in the hierarchy is applied to a nonlinear transformation on the input to create a statistical model as predicted output. The learning iterations continue until the output has reached an acceptable level of accuracy.

Deep learning has been growing at an incredible pace in 2018 and will continue to grow years beyond 2018. The recent trends in deep learning include supervised learning, reinforcement learning, networks powered by memory, edge computing and simulation environments. The global deep learning market size is growing at giant proportions and sensing opportunities. Deep learning is witnessing investments from Governments and business enterprises. These investments and improvements are driving business growth and revenues.

A company which stands tall in deploying deep learning on the development of 3D computer vision solutions in the next generation of gaming platforms, digital signage, and virtual reality is 3DiVi, Inc.

3DiVi is a revenue-generating developer of Human-centric AI algorithms, covering face recognition, skeletal tracking, spatial understanding and object recognition. Its key offerings include Nuitrack, the industry-leading 3D skeletal tracking middleware (Kinect replacement) which provides an alternative as a cross-platform skeletal tracking and gesture recognition solution for Android, Windows, Linux, and iOS platforms. Other product offerings are the Face SDK, a face recognition software, top-ranked in NIST Face Recognition Vendor Test 2017, and TVico – Interactive Android Computer. 3DiVi, founded in 2011, is headquartered in California with its R&D office in Chelyabinsk, Russia.

 

The Mission to Deliver Human Cognition Services

3DiVi Inc.is aimed to provide human cognition services to the edge. The company’s strategy is to build best-in-class human-centric AI platform and productivity services for an intelligent cloud and an intelligent edge. To deliver its mission, 3DiVi provides complete modules that can integrate face recognition into access systems, time and attendance systems and video surveillance systems of banks, stores, shopping and office centers, and other facilities requiring ready-made, easily scalable solutions.

 

Charismatic Leadership

 3DiVi’s Co-founders Paul Zaytsev, Andrey Valik and Dmitry Morozov have a synergetic combination of skills. Pavel is a successful entrepreneur, founder of Papillon, one of the leading biometric companies with $20+ million in annual sales. Andrey has 15+ years of experience in hardware/software development. Dmitry holds two US patents and an MBA from Owen Graduate School of Management that facilitates international marketing. Paul and Dmitry have both graduated from the leading Russian Tech University – Moscow Institute of Physics and Technology (MIPT).

 

A Proactive Approach into Disruptive Technology

3DiVi Inc. not only has a set of industry-leading human cognition algorithms but deploys a proprietary toolset for Synthetic data generator that drastically reduces the cost of developing custom object detectors. The company is a pioneer in Face and Body tracking solutions which bring a lot of value to multiple verticals requiring face-to-face customer interactions.

Travel, shopping outlets, healthcare, finance, education – all of them may benefit from customer recognition. According to CITA forecast, in 2020, 48% of airports and 43% of all airlines plan to implement face recognition into their business processes. A similar opportunity exists for offline loyalty programs – by recognizing in-store customers, businesses may streamline checkout. At the same time, with GDPR personal data privacy regulations coming into effect on May, 25th, 2018, new solution which guides the use of personal data is required.

 

Strategic Partnerships Enabling Innovations

3DiVi’s algorithmic products for face and body recognition are well known in the industry for unparalleled quality and performance. Among its current clients are Intel, LG, Orbbec (the leading manufacturer of 3D sensors in Asia) and Orange. The company focus on emerging areas of AR, Smart Home, IoT, Smart cars and robotics.

 

Awards and Media Mentions

3DiVi Face Recognition Algorithm is currently top-ranked in NIST FVRT 2017. The company has completed a successful Indiegogo campaign in 2016 (VicoVR Body Tracking Controller). 3DiVi has been a regular participant to Augmented World Expo, the largest AR & VR event in the world showcasing its innovative products VicoVR and Face Recognizing AR Glasses.

 

Challenges Along the Way

Dmitry said that a major concern for the company was talent staffing, the ability to attract, retain the best engineering talent in deep learning. 3DiVi solved it via its 3DiVi Computer Vision School, which was established back in 2014 in partnership with several local universities to train and retain its valuable human resources.

 

Future Ahead

3DiVi is poised to grow, challenging the market giants in human cognition offerings. “As the industry moves from SAAS to Edge Computing, we see an opportunity for smaller companies to challenge Microsoft Azure and Amazon AWS human cognition offerings for AR, IoT, smart home, smart car, robotics, and digital health verticals,” mentions Dmitry.