Event Location: NSH 1507
Bio: Stella X. Yu got her Ph.D. from the School of Computer Science at
Carnegie Mellon University, where she studied robotics at the Robotics
Institute and vision science at the Center for the Neural Basis of
Cognition. She continued her computer vision research as a postdoc at
the UC Berkeley’s Department of Computer Science. Since she joined
Boston College, Dr. Yu has been developing an interdisciplinary
curriculum and research agenda around art and visual perception. She
received the NSF CAREER award in 2007 on the topic of Art and Vision:
Scene Layout from Pictorial Cues.

Abstract: Decomposing a signal into multiple scales brings computational
convenience and speedup, as demonstrated by multi-grid methods in linear
algebra, wavelets for image compression, and Laplacian pyramids in
computer vision. However, scale also plays an important role in
revealing multiple visual contexts, which when examined as a gestalt
provide new insights into visual computation.

I will present 5 human and computer vision studies which explore 1) the
power of scale invariance in image segmentation, 2) the computational
challenge of recovering scale in deformable object matching, 3) the
distinctive processing characteristics of scale in human vision, 4) the
progression of featural scale in the categorization of spatial layout,
and 5) finally the integration of mixed scales in modeling brightness
perception without parsing image intensity into depths, surfaces, and
illuminations.