Computational and Mathematical Models in Vision (MODVIS)
Wednesday, May 11 – Friday, May 13
Organizers: Jeff Mulligan, NASA Ames Research Center; Zyg Pizlo, Purdue University; Anne Sereno, U. Texas Health Science Center at Houston; Qasim Zaidi, SUNY College of Optometry
The 5th VSS satellite workshop on Computational and Mathematical Models in Vision (MODVIS) will feature extended oral presentations on quantitative modeling of a variety of visual processes. A keynote address will be given by Prof. Steve Zucker (Yale) on Thursday afternoon.
The registration fee is $100 for regular participants, $50 for students.
WorldViz Virtual Reality Workshop
Saturday, May 14, 12:45 – 2:00 pm, Talk Room 2
Organizer: Matthias Pusch, WorldViz
Virtual Reality gets a lot of press lately, and releases about new hardware and software are coming out pretty much daily. This session will give an overview of recent trends and new technologies and discuss their use cases and limitations for research applications.
“Individual Differences in Vision” Brown Bag Lunch
Sunday, May 15, 12:45 – 2:00 pm, Talk Room 2
Organizers: Hrag Pailian1 and Jeremy Wilmer2; 1Harvard University, 2Wellesley College
Third biennial IDV Brown Bag: A whirlwind tour of the breadth of individual differences related work currently being conducted by VSS members. The event features a series of “micro-talks” where speakers give 2-minute presentations on cutting edge research that ranges across a wide variety of content areas. Bring your lunch, meet fellow researchers, and experience the power of the individual differences approach to vision!
Are there donuts in vision? Neural computation of global image configuration by a circular receptive field.
Monday, May 16, 2:00 – 4:00 pm, Sawgrass
Organizers: Naoki Kogo1, Bart Machilsen1, Michele Cox2, Vicky Froyen1; 1 Laboratory of Experimental Psychology, University of Leuven, Leuven, Belgium, 2Department of Psychology, Vanderbilt University, USA
The global visual context influences the processing of local image information in the visual system. Ample evidence for this claim comes from neurophysiological and psychophysical studies. This special interest/discussion meeting will explore plausible neural mechanisms that reflect the global configuration of an image, and the role of such a mechanism in Gestalt-like phenomena of figure-ground organization and shape representation. These phenomena demonstrate that long-range neuronal interactions modulate the coding of local image features in early visual areas. Border-ownership selective neurons, for instance, modulate their firing rate in response to the figural side of a boundary, which can only be derived from the global image context (1). To explain this mechanism, the existence of a “grouping cell” in a recurrent neural circuit has been suggested (2). From their functional role in the computation of border ownership some structural properties of these putative grouping cells have been proposed, such as a donut-shaped receptive field. The concept of a grouping cell has relevance for various lines of research: computational neuroscience, neurophysiology, and experimental psychology. This satellite event wants to facilitate cross-talk between these disciplines by approaching this putative grouping cell from different angles. For instance, the collective activities of grouping cells can result in a “medial axis” representation, similar to the skeleton-type of shape coding developed in computer vision science (3). Also, the pattern of oriented edges in natural images has revealed a co-circular organization of luminance boundaries (4), likely due to the presence of object boundaries (5). Neurophysiological data further hint at the existence of computational mechanisms to encode global configurations: Skeleton-like neural signals have been measured in V4 (6), an area where donut-shaped receptive fields have been found (7). Grouping-cell activity has also been reported for neurons aligned to the center of an illusory surface (8). This finding is in line with a recent computational model linking the coding of border-ownership to illusory surface perception (9). In this context, the neural representation of occluded shapes can help understand the neural coding of shape and depth order (10).
Taken together, the integration of approximately iso-distant edge information appears sensible given the co-circular pattern of edges in natural images and given the neurophysiological measurements and models described above. It remains to be seen whether this integrative mechanism could indeed be embedded in grouping cells with a donut-shaped receptive field. In this satellite event, the discussants from multidisciplinary backgrounds will examine the biological plausibility of this idea and discuss possible alternatives for the neural computation of global configurations in images.
Females of Vision et al. (FoVea) Meeting
Monday May 16, 5:00 – 6:00 pm, Breck Deck North
Organizers: Diane Beck, University of Illinois; Mary Peterson, University of Arizona; Karen Schloss, Brown University; Allison Sekuler, McMaster University.
VSS women and men of all ages and stages are invited to join us for the founding gathering of Females of Vision et al. (FoVea). We’ll discuss how, collectively, we can address issues to enhance participation in and success of women in vision science. Beverages will be provided on a first-come first-served basis until our $ runs out.