Natural Sciences and Engineering Research Council of Canada
Symbol of the Government of Canada

Common menu bar links

Past Winner
2007 E.W.R. Steacie Memorial Fellowship

Nikolaus Troje

Psychology

Queen's University


Nikolaus Troje
Nikolaus Troje

To call Nikolaus Troje's background eclectic would be an understatement. He teaches in both the Department of Psychology and the School of Computing at Queen's University, yet his academic training focused on biology, animal physiology and biopsychology. Throw in a few years spent farming and a liberal dose of physics and mathematics, and you get a researcher who has earned a reputation as the world foremost expert on analysing how humans and animals perceive and process biological motion – the movement of others.

The study of visual systems has been a constant focus of Dr. Troje's academic career. His early work involved studying vision in insects. He then moved to human face recognition, and was struck immediately by humans' ability to identify other faces with extremely subtle clues. So subtle, in fact, that despite advances in biometrics – the technology that uses data from faces and other parts of the body to uniquely identify people – the brain is still far more capable than the computer.

About seven years ago he became fascinated by the still more subtle visual clues provided by the way a person moves and the ease with which the human brain makes sense of them. Imagine a person moving through a dark room, their form outlined by a series of small point lights. Using just the information from the motion of the lights, another person virtually instantly knows not just that the figure is human, but can predict with a high degree of accuracy such other details as gender, age and emotional state.

When it comes to the animal kingdom, the ability to process visual information is critical to survival. An animal needs to be able to almost instantly identify that another creature is present, then equally quickly decide (based on the available information) whether that creature is friend, foe or indifferent, and react accordingly. Humans retain aspects of that innate ability, but add elements that are learned specifically from their cultures.

Groundbreaking experiments in the 1970s first demonstrated how finely tuned the brain's perceptions are with respect to biological motion, but the field remained underdeveloped for many years. The advent within the last two decades of more sophisticated technology for capturing motion and breaking it down into useful pieces of data helped revive it. Today's motion capture process uses high-speed cameras in conjunction with reflective dots to record three-dimensional records of the movement of people's joints.

Given the vast amount of information "encoded" in human movement, the challenge was to figure out how to turn that into data that a computer can crunch in a meaningful way, followed by understanding how the human visual system captures and uses this information.

Dr. Troje has made significant contributions to vision research and cognitive neuroscience by using motion capture technology in creative, new ways and helping develop the analytical tools needed to make sense of the data. The results have led to important insights into how people process motion.

One of his conclusions is that the old view of biological motion perception as a single phenomenon must go by the wayside. Instead, it should be regarded as a number of processes, each governed by a different area of the brain. In this tag-team approach, signals may start in a portion of the brain that can answer the question, "Is there something out there?" then get sent to other areas that answer, "What is it?" "What is it doing?" and "How should I react?"

Each of these processes likely originated at a different stage of evolution. The very basic aspects, which are associated with "fight or flight" reflexes, are considered very old and would therefore be found in a wide range of organisms. Other aspects, such as the ability to discern emotional states, are thought to be a more recent phenomenon and will therefore be found only in more developed animals and in humans.

Part of Dr. Troje's work during his Steacie Fellowship will involve studying the more innate aspects of motion detection in infants. Scientists already know that newborn chicks, for example, are hardwired to "imprint" on another creature, but that characteristic is guided by innate preferences based on very simple cues. Such impulses may also be present in humans.

Given Dr. Troje's varied background, it comes as no surprise that studying motion is a very multidisciplinary effort. Fields such as psychology, neuroscience, computer science and mathematics all play an important role.

In addition to collaborating with partners in different disciplines, he makes use of the fact that his Web site attracts an average of 10,000 visitors per day by running on-line experiments and by providing demonstrations that help disseminate his research.

Dr. Troje's work can be applied in a variety of ways, including helping the computer graphics industry create more lifelike characters by making their movements more natural. As movies and computer games evolve and computer-generated characters look more and more realistic, the demand for realistic-looking movement increases as well.

In a more serious vein, and of greater interest to Dr. Troje, is another area that he will focus on during his NSERC E.W.R. Steacie Memorial Fellowship – studying the potential implications of his research for people whose ability to process information from biological motion is impaired, such as people with autism. "They really suffer from the lack of social abilities that probably go back to the very basic sensory impairments in the area of biological motion perception," comments Dr. Troje.

"The goal is really to clearly identify what these building blocks of biological motion perception are," he says. By doing that, he hopes to develop tests that can distinguish between the various mechanisms, and from there to find ways to treat individual problems.