Breakthrough Information Technology Center
Our team of the World’s best programmers, designers has developed a Facial Recognition (FR) application that can be installed on mobile phone, tablet or Smart Glasses, with the main functionality of face detection and recognition with high quality. The idea is to use images from camera inside of smart-glasses, mobile phone or tablet to detect and recognize the faces of people around the operator (who wears the glasses). It is planned to move recognition inside of the glasses or other embedded device. Based on 15 years of research in the Computer Vision and Machine Learning, we have developed the unique and advanced algorithms and methods that allow us to detect and recognize faces, road signs, cars and people at a longer range and with the higher accuracy. We are patenting our own algorithm which is the hybrid approaches and methods based on statistical models.
Our Plan
Our goal is the development and gradual improvement plan for SmartGlassesFR products. We split our work into following main directions running in parallel:
-
To evaluate existing recognition libraries, like 3DiVi, develop application for smart glasses – application that captures images, selects faces (or just scales image before sending and makes some initial steps) and sends to server, after recognition receives information about faces and shows it. This application should do the following: Video capturing, Image scaling, Communication with server and Results visualization.
-
To develop SW for recognition server – server SW which receives images from smart glasses or smart phones and detects and recognizes faces, then sends data back to glasses/mobile: User API, Training on faces dataset, Communication with glasses.
-
To develop proprietary recognition engine to replace 3DiVI engine with our own. Evaluate existing algorithms, Select promising approach, Develop own recognition library.
-
Port recognition to embedded platform – evaluate possibility to port part of logic into glasses or smartphone, together with remote control/update engine: Port own library on smartphone, Develop remote control engine, Prototypes construction and testing – building of prototypes together for algorithms in-field evaluation.
We can also detect and recognize the following objects:
Cars
Road Signs
Some Animals
Lane Detection
Number Plates
USE CASES
Every three seconds someone in the world gets Alzheimer's disease. In 2015, there were 46.8 million patients, and by 2025 their number will exceed 131 million (Alzheimer’s Disease International). The same applies to people with a rare blood group, and so on. Millions of people will add themselves and their relatives to our database to provide important data about their loved ones for their safety.
If a person gets lost, may be unconscious or in any other emergency, any passerby, having our application, can look at the lost person through the phone or smart glasses and find out the name, special conditions (blood type or illness) and the contact number of the person who can be contacted in case of emergency!
If a person gets lost, may be unconscious or in any other emergency, any passerby, having our application, can look at the lost person through the phone or smart glasses and find out the name, special conditions (blood type or illness) and the contact number of the person who can be contacted in case of emergency!
Although most people do not recognize, 90% of families have lost children at least once. Children are lost more than 5,000 times a day in all public places, such as beaches, amusement parks, fairs and airports.
When people, for example are buying a ticket to Amusement Park, the child is immediately entered into the database with one click of the camera.
If a child is lost (statistically, it is harder to find parents, after finding a lost child ) any security guard, by pointing the phone or smart glasses at the child, will see his/her parents' contact numbers.
