NASHVILLE, Tenn. - Picture a marching band on the football field during halftime. With precise steps, some band members turn and march in a different direction than the rest, spelling out a school's letters or a mascot's shape. Scientists have known for years that to perceive figures against such a busy background, human vision uses color, brightness and direction of motion. But startling new findings reported by Vanderbilt researcher Randolph Blake in the May 14 issue of Science magazine indicate that the human brain can also use the precise timing of subtle visual changes to group elements into objects.
"People can compute changes in the direction of movement of many distinct parts of a scene at once, and can tell from the timing of these changes which parts belong to the same object," said Blake, a professor of psychology and investigator for Vanderbilt's John F. Kennedy Center for Research on Education and Human Development, who conducted his recent research with Vanderbilt graduate student Sang-Hun Lee. "This is the most convincing demonstration yet that the human visual system can use timing information alone to form coherent objects from features of a scene that change at the same time."
For their study, "Visual Form Created Solely from Temporal Structure," Blake and Lee created on a computer video screen a dense array of pinwheel-like objects, each spinning clockwise or counterclockwise. Each pinwheel changed its direction of rotation at random time intervals. When a cluster of pinwheels changed their directions at the same time, they appeared to form a coherent shape that stood out from the remaining pinwheels.
"Viewers were able to distinguish the synchronized group of pinwheels from the rest. This ability implies that we can compute changes in direction in the movement of many distinct parts of a scene and then identify the boundaries where the timing of these changes differs significantly," explained Blake. "In one respect, these are highly artificial visual displays," he said of the experiments. "It's hard to imagine an everyday situation where we would encounter an object whose presence would be defined solely by temporal coincidence. However, as we move about in our environments and look at objects that are perhaps themselves moving, temporal coincidence is inevitably part of the optical input that we must rely on for making sense of those objects and events."
Blake conducts his research - an interface of psychology and neuroscience - at the Vanderbilt Vision Research Center, founded in 1989 with a grant from the National Eye Institute. "I'm investigating human visual perception," he explained. "Vision - and in fact all of perception - happens quickly and effortlessly, yet the steps involved are extremely complicated. This complexity is revealed when engineers try to build machines that see. It turns out to be a huge technical challenge that has not been mastered. Yet our brains accomplish visual perception every waking moment of our lives."
Blake, who served as chair of Vanderbilt's psychology department from 1988 to 1996 and is a fellow of the American Association for the Advancement of Science and of the American Psychological Society, believes his work may advance understanding about the coding scheme that the brain normally uses in ordinary viewing situations.
"At the moment, there are strong opinions about the possible role of neural timing in perceptual grouping," Blake said. "This is an issue that Sang-Hun Lee and I prefer to leave to the neurophysiologists. But we believe our novel displays, and people's remarkable ability to perceive form based on temporal synchrony, may be useful in resolving the issue."
For more news about Vanderbilt, visit the News Service home page on the Internet at http://www.vanderbilt.edu/News. To view this Science story on May 14, see: http://www.sciencemag.org. Additional information on Blake's work can be found at: http://www.psy.vanderbilt.edu/faculty/blake/DEMOS/TS/TS.html
The above story is based on materials provided by Vanderbilt University. Note: Materials may be edited for content and length.
Cite This Page: