Adding dimension to data: Daniel Keefe

Movement, pattern, shape: For a scientist, each set of data can be as unique and inspiring as a work of art. McKnight Land-Grant professor Daniel Keefe paints a picture that sheds new light on massive data sets.

Keefe uses modeling and computer visualization to interact with information in three dimensions. His insight on technical and visual modeling is in high demand by researchers who want to get more from their data.

Keefe is collaborating with faculty from across the University of Minnesota — designing medical devices, understanding biomechanics data, designing tools to train surgeons, and building virtual reconstructions of ancient Greece, to name just a few projects.

“If we come up with the right visualization and data analysis tools, then we can make it possible for scientists and engineers to really take advantage of the data they collect,” says Keefe.

While the work of comparing hundreds of variations of a given data set in a single visual space remains daunting, Keefe is up for the challenge. For example, he intends to map the movement of a bat’s wing by comparing the movement of hundreds of bats in multiple dimensions.

“It becomes a problem which has both a challenging computational portion to it, but then it also has a challenging visual design problem. That’s a really fascinating question.”

Art meets science
Keefe’s mother is a printmaker, and art drove his curiosity from an early age. He was eager to contribute to the prevalent link between art, science and computer graphics in research. As an undergraduate, he spent a summer with NASA, visualizing air flow around landing gear.

“At the time, I was taking a lot of art classes, which highlighted for me a really nice overlap between technology and visual settings. That got me hooked.”

Keefe encourages his computer science students to take advantage of art classes at the university, emphasizing that a familiarity with design is imperative in visual computing. His research partners have included artists and designers from the Minneapolis College of Art and Design and the Rhode Island School of Design.

He says the most gratifying part of teaching is the transformation he sees in each student, when they begin demonstrating the concepts that excite them.

“You feel as if you’ve really taught them something and now they are beginning to pass it on — it’s really clicking.”

What’s next?
When asked to list the newest developments in visual computing research, Keefe points to the following trends:

Borrowing concepts from gaming: Only a few years ago, scientists were spending millions to equip their labs with the visualization technology now available to gamers, e.g. Kinect for Xbox 360 and PlayStation Move. Keefe says scientists are beginning to borrow and modify these tools to create virtual reality environments, in any space, for a marginal cost.

Scaling up: Visual computing is becoming more sophisticated, and better able to accommodate hundreds, sometimes thousands, of data sets. The visual computing field will refine a process for representing a series of data sets in the same space, to help researchers better identify meaningful trends across data.

Photos by Andria Peters

Republished with permission from Research @ the U of M, a publication of the Office of the Vice President for Research.