Machine Learning Use Case

The Challenge

To create a facial recognition system that automatically identifies and tracks an individual in a crowded environment. A system that is user-friendly, scalable and uses customisable facial recognition algorithms in order to operate over IP and be well suited for integration within Cloud solutions. The identification process needs to be enabled by comparing the facial features extracted from a digital image or video, with those previously stored in a facial database.

The Solution

Showcasing high-tech identification technology merged with advanced facial tracking, we have developed a custom facial recognition software designed to fortify safety and security in places where large numbers of people congregate, such as football stadiums.

By building a tailored algorithm that required high-speed bandwidth to support optimum tracking, our facial recognition system was developed to demonstrate a potential use-case for 5G. Designed to be stationed as both a proactive and preventive measure at points of access, this technology is well suited for public locations such as airports and stadiums, where there is mass transit of people. The algorithm can be tailored to purpose built databases, and provides a flexible solution for multiple settings.


1. Multiple-matching facial detection tailored for crowds

2. Rapid processing time with minimal lag

3. Peak recognition rate of 96%

4. Algorithm adaptive to vantage points and facial changes


Despite the evolving surveillance industry and increasing requirements for biometric technologies being the major drivers of growth behind the market, the further potential of facial recognition lies in the untouched, yet people orientated, areas of administration, advertising, automotive, media, tourism, and medical fields. Facial recognition can bring a new dimension to IoT devices, enable automated social media tagging and personalised advertising.

© 2019. All rights Reserved
roomone | Terms | Privacy