Professor Hassan Ugail - Research in Visual Computing at Bradford

Computer graphics is a field that is approximately 40 years old. Over this period the field has developed enormously, to the point where the techniques and technology of computer graphics are ubiquitous in both the home and workplace and underpin fields as diverse as entertainment and e-Research.

Yet despite these rapid technological advances, the shape of computer graphics as a subject does not yet resemble a core discipline in science or engineering. A group of academic researchers at Bradford are working in this area under the theme of Visual Computing. Their task is to enable computers to efficiently perceive process and understand visual data such as images, videos and complex 3D scenes.

Most of us when we hear the terms such as "computer graphics" and "visual computing" associate them with the entertainment industry such as computer gaming and films. The fact of the matter is that this is an area which has a rich variety of applications beyond films and computer gaming. For example, let us take the case of human facial animation, i.e. generating human facial expressions (facial stimuli) on the computer. Human facial stimuli are used to help understand disorders such as Alzheimer's and Parkinson's disease. It can be used for teaching social skills and has been found to be successful in improving recognition of emotions in autistic individuals. Facial stimuli can help to better understand mental illness, e.g. to test the effects of emotion recognition in sufferers of schizophrenia and depression.

In 1999, while I was still a PhD student I came across a hypothesis called the "uncanny valley" which was developed by a Japanese researcher called Mori who was working on the design of robots. Mori's hypothesis states that as a robot is made more humanlike in its appearance and motion, the emotional response from a human being to the robot will become increasingly positive and empathic, until a point is reached beyond which the response quickly becomes that of strong revulsion. However, as the appearance and motion continue to become less distinguishable from a human being, the emotional response becomes positive again and approaches the same level of human empathy.

This so-called uncanny valley effect is actually not just limited to robots. In fact it is applicable to any type of human-like object, e.g. dolls, masks, avatars in virtual reality and characters in computer graphics games and movies. For example, while we might empathise with a cartoon-type character (e.g. Bambi) we find it more difficult to relate to facial animations which aim to mimic real humans (e.g. The Polar Express) – in fact, our reaction can be negative.
Much of the current research in computer graphics relates to the generation of realistic scenes by means of identifying ways of efficiently dealing with the uncanny valley effect.

Practitioners (such as those working in the movie-making industry) use a combination of ad-hoc tools to achieve reasonable results, and of course with a lot of effort and resources.

However, in order to create a truly realistic computer graphics, one needs to take into account many other factors (beyond the uncanny valley effect), and this requires not only the knowledge in the field of computer science but also other fields such as mathematics, engineering and psychology. Computer graphics as a subject is relatively young and very dynamic. Though it does have some of the more basic engineering features and much of it derives from scientific principles, there has been little consolidation of the theory.

In recent years much published work has bee strongly driven by secondary technology developments, such as computer power and graphics cards; or by big-budget but highly specific applications, such as movies.

Understandable though this is, it has the effect that the corpus of knowledge is growing faster than the opportunity to consolidate, amplify and unify that knowledge. In turn this makes it difficult to discern new experimental work to test or challenge the theory.

A mature science discipline is characterised by a corpus of knowledge leading to a range of theory and then to experiment to test that theory. It makes progress by identifying research problems needing solutions. These may arise from where the theory does not match the experimental results or where no experiment has yet been devised to test a certain aspect of the theory. More rarely, major theoretical breakthroughs may occur, detached from known experimental lines. Typically, theories within a given discipline converge to common principles and less-diverse assumptions. The explanatory power of the unified theory is accordingly magnified.

For example, a mature engineering discipline takes the relevant science and established practice to predict the behaviour of its materials and the ways they can be combined or formed to make useful artefacts. It explores and quantifies the limits of materials and technologies. It investigates, by combining them in novel ways, to produce enhanced qualities or to permit their use in more extreme environments. Practice may drive early exploration but theory is always sought to explain the experimental results and so extend the range of application. Computer graphics as a subject is relatively young. Though it may have some of the more basic scientific features, there is little in terms of a theory.

In many ways I view research in this area to be a re-working of known laws of nature such as Newton's laws of motion and Einstein's theory of relativity in a suitable form applicable to a virtual world where physical matter does not exist. Having access to a "theory" of computer graphics will enable one to create truly realistic computer graphics which will have a major impact not just on the entertainment industry (e.g. for movies and video games), but also for understanding and treating a wide range of psychological and neurological disorders, such as Parkinson's, Alzheimer's, autism, schizophrenia and depression. This will indeed require serious effort from researchers and practitioners in order to combine knowledge from various fields, particularly from applied mathematics, engineering, core computer sciences, as well as psychology.

Much research and knowledge transfer work in this field is currently being undertaken at Bradford. A core group of researchers based at the School of Computing, Informatics and Media, School of Life Sciences and School of Engineering, Design and Technology are working on related research. Our research is publicly funded by leading research grant-awarding bodies such as the Engineering and Physical Sciences Research Council (EPSRC).

A notable research project we are currently engaged in includes the development of a "passive lie detector" for border control applications. In partnership with the University of Aberystwyth, UK Border Agency and QinetiQ we are developing a computer system to detect human guilt by means of analysing human facial expressions both in the visual and thermal domain. The system will link human emotions and physiological processes such as blood flow, eye movement patterns and pupil dilation and body language. The potential application for this system we have in mind is to profile people as they pass through border controls to help security agencies identify smugglers.

Of course we also have been working in the area of computer graphics applicable to computer games. Our recent research in computer graphics has led us to develop efficient ways of compressing 3D data especially those used in computer games. This has led us to the formation of a University spin-out company Tangentix Ltd with substantial funding from venture capitalists through Enterprise Ventures and South Yorkshire Investment Fund. Through Tangentix technology we are currently developing a platform for digital distribution of computer games content which has a global business market of $8bn.

As a further development we have recently received inward investment from the University which has enabled us to establish a centre of excellence here in Bradford called the "Centre for Visual Computing". This new centre brings together world-leading research in areas such as digital imaging and metrology, visualisation and human visual perception. The centre aims to be the partner of choice for private sector companies and public sector organisations operating in the global digital space and who are seeking digital solutions.

With the enormous potential this field has to offer and with these recent new developments, this is truly an exciting period for Bradford. After all, we are working within virtual worlds and hence imagination is the only limit!