Mark Everingham, BSc, PhD
Dr Mark Everingham, Lecturer in the School of Computing and one of the leading computer scientists of his generation, died on 29 June 2012.
Mark Everingham read Computer Science at the University of Manchester. He excelled at his studies throughout his time as an undergraduate and his academic performance earned him the Addison-Wesley, Professor’s and ICL Prizes. In 1995, he graduated with First Class Honours and the Williams-Kilburn Medal for exceptional achievement in undergraduate Computer Science. During and after completion of his undergraduate degree, Mark also undertook research in the Bristol Eye Hospital, developing software for an automated electrodiagnostic system based on EOG (electrooculography, a technique for measuring the resting potential of the retina) and EEG (electroencephalography, a technique for recording brain activity) measurements.
Mark began his PhD studies at the University of Bristol in 1997. His research, on methods and metrics for image classification with application to a mobility aid for those with poor sight, was concerned with object recognition in computer vision. Computer vision is the branch of Computer Science that aims to replicate computationally the human capacity to use the eyes and brain to see and visually sense the physical world. Object recognition in computer vision refers to the demanding challenge of being able to locate a particular object in an image or video sequence. Mark was awarded his PhD in 2002 and in the same year delivered a paper based on his research to the 7th European Conference on Computer Vision. Presented with his hallmark wit and vigour, the paper, with its description of a valuable new methodology for future research into object recognition, proved to be one of the highlights of the conference.
Following completion of his PhD, Mark spent some months as a researcher for a biotechnology company. Here, he developed computer vision techniques for the segmentation and measurement of cells in fluorescence microscopy images. In November 2002, he was appointed as a post-doctoral research fellow in the Department of Engineering Science at the University of Oxford, working with Professor Andrew Zisserman, FRS, on a European-funded project on Cognitive Vision Systems (CogViSys). The CogViSys project was part of an ambitious long-term research programme directed at bringing a cognitive level of understanding to video material: enabling a blind person to identify a specific individual in a video, for example. Drawing on his extensive knowledge of the machine-learning literature, Mark made a very substantial contribution to the project. Successfully combining existing approaches into a unique new approach, his work led to a very much improved face detection algorithm and was successfully presented at both the International Conference on Image and Video Retrieval (CIVR) and the IEEE International Conference on Computer Vision (ICCV) in 2005. When the CogViSys project came to an end, he continued his research in human face detection and activity understanding. In 2004 he was awarded a Fulford Junior Research Fellowship at Somerville College.
Whilst at Oxford, Mark also became joint organiser, in 2005, of the first PASCAL (Pattern Analysis, Statistical Modelling and Computational Learning) Visual Object Classes (VOC) Challenge. The challenge, in which research groups compete, was set up to stimulate research into visual object category recognition and detection, and to provide the vision and machine learning communities with standard datasets for facilitating and evaluating research. The challenge has continued to be organised on an annual basis since 2005 and its associated datasets have come to be accepted as the benchmark for object detection. Mark remained one of the leading lights in organising and co-ordinating the challenge. His dedication and enthusiasm, and above all his uncompromising insistence on the most rigorous experimental evaluation, have left an indelible imprint on this field of research. In consultation with Mark, it had been decided that 2012 should be the last year of the challenge. The final VOC competition and workshop will now be dedicated to the memory of Mark.
It was during his time at Oxford, too, that Mark began to develop an increasingly world-wide profile. This was shown in the growing number of invitations he received to review papers for international conferences and journals, and to serve on conference programme committees.
Mark came to Leeds in 2006, to take up a prestigious appointment as RCUK Academic Research Fellow in the School of Computing. Here, he built up a successful team of PhD students working in the area of object recognition and with them continued to publish in the major conferences in Computer Vision. He also started to develop a portfolio of externally-funded research, most recently in collaboration with Professor Richard Bowden at the University of Surrey and Professor Andrew Zisserman at the University of Oxford. In 2011, Mark became an Associate Editor of the IEEE Transactions on Pattern Recognition and Machine Intelligence, one of the leading journals in the field. On completion of his Fellowship in 2011 he was appointed Lecturer in the School. It was quickly evident that the outstanding rigour and dedication that underpinned his research also made Mark a gifted teacher. He was asked to develop a module for the School's new IT programme, material for which there was little precedent in the curriculum. Mark took on the challenge and produced material that, in its first outing, earned plaudits from the class. His approach was straightforward - to be outstanding in teaching, as in research, required developing a deep understanding of the problem to be solved, subjecting any idea to his own exacting standards, and only delivering material, be it a research paper or lecture, that was the best he could achieve. Research students whom Mark supervised or mentored would find their ideas subjected to the fiercest intellectual examination in review meetings, but these were never less that inspirational as a result.
Despite his relative youth, Mark Everingham was a computer scientist of very considerable achievement who would undoubtedly have gone on to become one of the dominant international figures in the field of computer vision. At all times, he set himself the highest of standards and expected nothing less from those around him. Yet he was an excellent colleague, generous with his counsel and always ready to acknowledge and encourage the work of others. The depth of respect among his peers for his intellectual qualities can be gauged by the number of them who exposed their planned papers and presentations to his relentlessly probing and comprehensive scrutiny, secure in the knowledge that if their work passed muster with such a stern and demanding critic, it was virtually guaranteed to be well received by the scientific community.
Mark is survived by his partner, Jelena.
The funeral has taken place. Jelena and Mark’s family also organised a gathering in Leeds earlier this month at which friends and colleagues had the opportunity to come together to celebrate Mark’s life and work.