For digital rights groups, facial recognition software might signal the beginning of the end when it comes to personal privacy and security. But on the other side of the debate, there is an argument that our faces could be the key to unlocking huge potential in the modern world.
Big Brother Watch chief executive Renate Samson and The Next Web reporter Matthew Hughes recently painted a bleak view of the dystopia facial recognition software is leading us into, with the imminent launch of the iPhone X at the helm of a ship that has already set sail.
This is far from the view held by Lyndon Smith PhD, who researches 3D vision systems at the Bristol Robotics Laboratory. Continuing the facial recognition software debate, he raises the question: when it comes to the ethical implications, is the technology really to blame?
Lyndon Smith PhD, Professor in Computer Simulation and Machine Vision at the University of the West of England (UWE), Bristol.
Face recognition has been introduced into various applications over the last few years, with vision systems being employed in airports and some retail businesses. However, up to now systems have employed conventional (i.e. 2D) images, similar to those captured by a digital camera. The problem with this approach is that it can easily be ‘spoofed’ by someone holding a photo of someone else in front of their face. This, in combination with the fact that recognition rates decrease when a scene changes by, for example, the sun coming out or additional people appearing in the image, has so far limited the applications of this technology. However, this is all about to change.
3D face recognition can tell the difference between a real person and a photograph and be made resilient to changes in background light. In the Centre for Machine Vision at the Bristol Robotics Laboratory (BRL), UWE, Bristol, we have been researching into 3D vision systems for over ten years. A recent project involved developing a system for use on the London Underground for eliminating the need for physical gates. This approach promises great benefits for reducing delays at currently congested stations such as London Victoria, and the project received significant media attention.
The public interest in this technology was confirmed by the publicity surrounding the launch of the iPhone X; however, this device has a limited 3D resolution capability. It can in fact be thought of as an Xbox v. 1 Kinect interfaced to a laptop and miniaturised – and the relatively low 3D resolution (173x173 pixels) will limit the reliability with which faces can be identified.
In contrast to this, our systems (which employ structured lighting) are only limited in resolution by the resolution of the camera employed – and cameras of 5MPix and above [on devices] are readily attainable at reasonable cost. This enables us to build very high-resolution systems that offer potential for recovering a ‘3D fingerprint’ from the face - thereby providing face recognition reliabilities approaching 100%. This effectively lets the tiger out of the cage and the potential applications are almost endless.
It is worth stressing however, that this technology is intended only to benefit people by making their lives easier and then only for persons who wish to employ it. Many people are, perhaps understandably, concerned about the ‘Big Brother’ potential of a capability for individuals to be recognised and tracked. However, whenever we walk down the street in any major UK city we are already being photographed hundreds of times – the important question is what then happens to this data. Advanced technologies provide advanced capabilities, which are intrinsically neither good nor evil; rather it is the decisions of lawmakers and policy-setters that ultimately decide whether the systems involved have a positive or negative impact on the quality of our lives.
If the technology is used correctly the potential positive benefits are immense. Imagine being able to access everything you need every day without the need to carry multiple cards around and to memorise PINs - imagine if presenting your face were the key to getting nearly everything you want in the modern world.
For more information on CMV 3D face recognition technology, please visit the Centre for Machine Vision website.
Read Part One of the debate, or watch the Lush Creative Showcase panel debate on the future of the open internet.