Microsoft Research Blog

The Microsoft Research blog provides in-depth views and perspectives from our researchers, scientists and engineers, plus announcements about noteworthy events, scholarships, and fellowships designed for academic and scientific communities.

Fitzgibbon’s Research Reaps Silver Dividend

June 28, 2013 | Posted by Microsoft Research Blog

Posted by Rob Knies

Andrew Fitzgibbon

Research success can be characterized in any number of ways. It might be by the cleverness of an algorithm. It might be by paper citations, or product contributions, or helping to develop disruptive technologies.

Sometimes, though, having a successful research career can be as simple as pursuing a path of lifelong learning. Just ask Andrew Fitzgibbon.

On June 27, Fitzgibbon, a principal researcher at Microsoft Research Cambridge, was announced as one of four winners of the Royal Academy of Engineering’s Silver Medal for 2013. The award recognizes outstanding and demonstrated personal contributions to British engineering, resulting in successful market exploitation, by an engineer with less than 22 years of full-time employment.

Fitzgibbon, of course, was delighted to hear the news.

“It’s fantastic to be recognized by the U.K.’s national academy for engineering for turning research ideas into practice,” he says. “We often talk about theory and practice as if there’s a one-way link: First you think of the theory, then you mangle it until it works in practice. I believe that only by considering its concrete impact can one develop really valuable theory.

“It’s in some sense easy to have a new idea, but only by grounding it in the real world can one be sure it’s a good idea.”

Note above the words “successful market exploitation.” The Silver Medal is not presented merely for great engineering. Recipients must have achieved significant commercial success in their fields and be recognized for advancing the cause of engineering in the United Kingdom.

Fitzgibbon, who joined Microsoft Research Cambridge in 2005, has scaled that mountain a couple of times. After a successful collegiate career that included an undergraduate degree at University College Cork and a Ph.D. in artificial intelligence at The University of Edinburgh, he held a Royal Society University Research Fellowship at the University of Oxford.

It was there that Fitzgibbon, who has published extensively in the field of computer vision, served as the primary engineering influence in converting academic research at Oxford into a commercial product, the automatic camera tracker boujou, which helps 3-D animators glue special effects to a live-action background. By studying a sequence of images, boujou can compute the3-D motion of the camera that took the sequence, and that lets animators insert 3-D objects into a scene without the need for virtual stop-motion techniques.

So effective is this process that it has become de rigueur for use in films and TV series that use special effects, including such landmark movie franchises as Lord of the Rings and Harry Potter.

That, in itself, would constitute a career highlight for most, but Fitzgibbon had more tricks up his sleeve. After joining Microsoft Research Cambridge, he contributed expertise in machine learning for human-motion capture. That became a core technology driving Kinect for Xbox 360, and he then instigated the construction of a pipeline for generating data for the machine learning algorithm at an unprecedented scale. In 2011, he and his colleagues won the Royal Academy of Engineering's MacRobert Award, one of the U.K.'s most prestigious engineering innovation prizes, for that technical contribution.

One man, two distinct marketplace contributions—but, as Fitzgibbon explains it, there is a direct line from one to the other.

“On boujou,” he says, “I was the core engineer, in the sense that I wrote code to implement the computer-vision algorithms, while others managed the user interface, software architecture, and testing. The latter role, testing, was carried out by the project manager—Andrew Stoddart, now at Sony—and was at least as important as the core engineering. Every day, we found new data to break our algorithms and new metrics to identify such breakages.

“From this experience, I learned the value of gathering extensive and challenging test data, which meant that when we started to work on Kinect, I immediately identified the gathering of test and training data as crucial to the project’s success. I saw that as my opportunity to contribute, while others worked on the core algorithms.”

The Royal Academy of Engineering, which promotes the engineering and technological welfare of the United Kingdom, established the Silver Medal in 1994. Interestingly, Andrew Blake—a renowned figure in computer vision who is a Microsoft distinguished scientist and laboratory director of Microsoft Research Cambridge—won the same award in 2006.

For Fitzgibbon, that connection, too, represents a learning experience.

“Since long before I joined Microsoft,” he says, “indeed, during my Ph.D. in the early ’90s, I used to write every paper guided by the mental exercise: ‘What would Andrew Blake think if he read this?’ That served to sharpen my arguments and to focus the writing.

“It is indeed an honour to join any group of which Blake is a member.”