http://www1.psych.purdue.edu/~zpizlo/

Dr. Zygmunt Pizlo is a Professor of Psychology at Purdue University. He received his Ph.D. degree in Electronic Engineering from the Institute of Electron Technology, Warsaw, Poland in 1982, and his Ph.D. degree in Psychology from the University of Maryland, College Park, MD in 1991. His research interests cover all aspects of vision, including shape, symmetry, depth, motion, binocular, color, figure-ground organization, 3D scene recovery, eye movements, image and video quality assessment, speed-accuracy tradeoff, as well as human motor control and problem solving. His psychophysical studies are always accompanied by mathematical and computational modeling: perspective and projective invariants, multiresolution and graph pyramids, regularization and Bayesian methods for solving inverse problems. In 2008 he published a book on human shape perception (MIT Press), which is the first monograph on this subject. His research has been supported by the National Science Foundation, Air Force Office of Scientific Research, US Department of Defense, US Department of Energy and Hewlett-Packard Company. He co-chaired interdisciplinary workshops on shape, on problem solving and tutorials on shape and on psychophysics. He was the president and the member of the executive board of the Society for Mathematical Psychology and he is the founding editor of the Journal of Problem Solving. He has published over 100 journal and conference papers and book chapters.

Abstract:

Perception of symmetry has been studied scientifically for the last 100 years, beginning with the seminal book by Mach. Mach pointed out that mirror symmetry is perceptually more salient than translational or rotational symmetry. Aesthetics and art provided the main motivation for the early studies, which concentrated on perception of symmetrical retinal images. The first use of symmetry as an a priori constraint in visual perception was described by the Gestalt Psychologists in the 1920s and 30s. They treated symmetry as representing a simplicity principle (Prägnanz), defined, informally, as economy of perceptual representation. Simplicity principle was responsible in Gestalt theory for perceptual organization, in which spatially global features took precedence over spatially local ones. During the Cognitive Revolution in the 1950s and 60s, symmetry was extended to the case of 3D stimuli and shown to be responsible for the topological and metric properties of the 3D shape percept produced by a single 2D retinal image. In the 1970s, symmetry perception was again studied separately from shape. The effect of retinal position, orientation, skew, as well as the effect of the amount of noise was tested in a set of parametric studies. The adoption of an inverse problem approach to vision, in conjunction with the progress in regularization and Bayesian methods for solving ill-posed inverse problems, brought, for the second time in history, symmetry and shape studies together. This research included the role of parts (geons) in perception of 3D objects, the application of Curie principle to shape perception, the relation between symmetry and binocular vision, and the dependence of symmetry constraint on other constraints, namely compactness, planarity and the non-degenerate view assumption. The tutorial will be concluded with a new definition of shape based on the concept of symmetry.

https://securewww.esat.kuleuven.be/psi/visics/people/?uid=1 and http://www.vision.ee.ethz.ch/members/get_member.cgi?lang=en&id=1

Luc Van Gool graduated in 82 as an electrical engineer at the University of Leuven in Belgium. He earned a PhD on invariants of planar contours. He is a full professor at both the University of Leuven and the ETH Zurich in Switzerland. He leads research groups working on computer vision at both universities. He is the co-founder of several spin-offs, namely Eyetronics, eSaturnus, kooaba, GeoAutomation, and Procedural.

Abstract:

Symmetric shapes, even when looked at obliquely, have invariants that irregular shapes do not have. So-called fixed structures play a key role in this. These are structures such as points or lines that remain fixed under the transformation expressing the symmetry. Under general perspective projection and for planar shapes, the fixed structures yield subgroups of the plane projectivities. We analyse some of the more interesting cases and derive example invariants for each.

http://www.cs.toronto.edu/~sven/

Sven Dickinson received the B.A.Sc. degree in Systems Design Engineering from the University of Waterloo, in 1983, and the M.S. and Ph.D. degrees in Computer Science from the University of Maryland, in 1988 and 1991, respectively. He is currently Professor and Chair of the Department of Computer Science at the University of Toronto, where he has also served as Acting Chair (2008-2009), Vice Chair (2003-2006), and Associate Professor (2000-2007). From 1995-2000, he was an Assistant Professor of Computer Science at Rutgers University, where he also held a joint appointment in the Rutgers Center for Cognitive Science (RuCCS) and membership in the Center for Discrete Mathematics and Theoretical Computer Science (DIMACS). From 1994-1995, he was a Research Assistant Professor in the Rutgers Center for Cognitive Science, and from 1991-1994, a Research Associate at the Artificial Intelligence Laboratory, University of Toronto. He has held affiliations with the MIT Media Laboratory (Visiting Scientist, 1992-1994), the University of Toronto (Visiting Assistant Professor, 1994-1997), and the Computer Vision Laboratory of the Center for Automation Research at the University of Maryland (Assistant Research Scientist, 1993-1994, Visiting Assistant Professor, 1994-1997). Prior to his academic career, he worked in the computer vision industry, designing image processing systems for Grinnell Systems Inc., San Jose, CA, 1983-1984, and optical character recognition systems for DEST, Inc., Milpitas, CA, 1984-1985.

Abstract:

Perceptual grouping played a prominent role in support of early object recognition systems, which typically took an input image and a database of shape models and identified which of the models was visible in the image. When the database was large, local features were not sufficiently distinctive to prune down the space of models to a manageable number that could be verified. However, when causally related shape features were grouped, using intermediate-level shape priors, e.g., cotermination, symmetry, and compactness, they formed effective shape indices and allowed databases to grow in size. In recent years, the recognition (categorization) community has focused on the object detection problem, in which the input image is searched for a specific target object. Since indexing is not required to select the target model, perceptual grouping is not required to construct a discriminative shape index; the existence of a much stronger object-level shape prior precludes the need for a weaker intermediate-level shape prior. As a result, perceptual grouping activity at our major conferences has diminished. However, there are clear signs that the recognition community is moving from appearance back to shape, and from detection back to unexpected object recognition. Shape-based perceptual grouping will play a critical role in facilitating this transition. But while causally related features must be grouped, they also need to be abstracted before they can be matched to categorical models. In this talk, I will focus on symmetry as a powerful regularity for decomposing a shape into parts. I will review a number of symmetric parts-based shape representations, including shock graphs, bone graphs, blobs/ridges, top-points, and medial surfaces, and show how they can support effective object indexing and recognition. I will also discuss the extent to which these representations provide shape abstraction, which is critical in order to handle the within-class shape variation that's required for object categorization.