There has been a great deal of work on face recognition technologies over at least the last 35 years including some on video based recognition. In 16 years of research we have implemented and evaluated many state‐of‐the‐art systems, but almost all methods we have tested to date failed miserably when tested on uncontrolled low resolution image probes against uncontrolled low quality face galleries. Formal benchmarking on passport quality images will often yield impressive recognition rates with virtually zero errors. Yet everyone with any experience in biometrics knows that such performance is simply unattainable in the field without enormously expensive image capture equipment. Traditionally obtaining good recognition rates is all about getting the image capture conditions absolutely perfect — and achieving this in the field is incredibly expensive.

In this presentation I will describe our biometrics and surveillance research work on a transcontinental surveillance project currently running with multi-camera face recognition appliance nodes in Australia, UK, and Brazil. These systems runs securely over the internet with edge processing to massively reduce bandwidth requirements and to improve corporate privacy. There is no requirement for an expensive dedicated fibre network to connect all the high speed cameras - indeed wireless and mobile connectivity is often a viable option. The cloud-based incident management backbone is accessible to all users from anywhere in the world. Along the way I will discuss the basics of robust CCTV-based video face recognition and the huge technical challenges of simultaneous pose, expression, illumination, obscuration, and motion blur compensation. I will also discuss quite recent work on robust face detection, landmarking, and tracking which enables our systems to work on a crowd of people walking quickly past the cameras. There will be live demonstrations including of mobile and possibly wearable face recognition apps.