keynote

Keynote Speakers

Keynote Speaker 1 - Martin S. Banks (Mon. June 10. 13:30 - 14:30)
Title: The Limits of Human Stereo in Space and Time
Abstract:
  In this talk, I will discuss how the visual system determines the displacement of an image patch in one eye that yields the highest correlation with a patch in the other eye. This occurs in visual cortex after spatio-temporal filtering of the monocular signals. The computation of inter-ocular correlation after filtering imposes fundamental bounds on the disparities that can support depth perception. I will also analyze those bounds experimentally by measuring humans' ability to see disparity variation for various spatial and temporal frequencies. Finally, I’ll describe the important insight into the properties of neural mechanisms underlying the perception of depth from disparity.



Martin S. Banks received the B.A. degree in psychology from Occidental College, Los Angeles, California, in 1970, the M.S. degree in experimental psychology from University of California, San Diego, in 1973, and the Ph.D. degree in developmental psychology from the University of Minnesota, Twin cities, Georgia, in 1976. In 1984, he joined the faculty of the Optometry, University of California at Berkeley, where he is currently a Professor. From 1995 to 2002, he was an Chairman of the Vision Science Program with University of Callifornia at Berkeley.

Keynote Speaker 2 - Alan Conrad Bovik (Tue. June 11. 09:00 – 10:00)
Title: Perspectives on Natural 3D Statistics
Abstract:
  In this talk I will discuss a variety of topics related to 3D images and the perception of them. Underlying my talk is the thesis that as with 2D images of the natural world, 3D images (stereo pairs or their associated disparity/depth maps) obey lawful statistical behavior that can be modeled. These models supply powerful statistical priors that can be used to define or regularize the solutions to a variety of visual problems, such as depth estimation from one or more images, stereo pair quality assessment (distortion severity assessment), and predicting the degree of discomfort felt when viewing stereo 3D (S3D) images. I’ll describe ongoing work in these areas and pose some general problems to be solved in the future.



Al Bovik is the Curry/Cullen Trust Endowed Chair Professor at The University of Texas at Austin. He has received a number of major awards from the IEEE Signal Processing Society, including: the Best Paper Award (2009); the Education Award (2007); the Technical Achievement Award (2005), the Distinguished Lecturer Award (2000); and the Meritorious Service Award (1998). He has also received the SPIE Technology Achievement Award in 2012, the IS&T Honorary Membership in 2013, and was named Imaging Scientist of the Year by IS&T/SPIE in 2011. He is the author or co-author of The Handbook of Image and Video Processing, Modern Image Quality Assessment, and two recent books, The Essential Guides to Image and Video Processing.

Keynote Speaker 3 - Marc Pollefeys (Tue. June 11. 13:20 - 14:20)
Title: Towards semantic 3D reconstruction from images
Abstract:
  In this talk I will present our recent work on 3D reconstruction from images. I will in particular focus on architectural scenes where specific priors can be used to obtain better and more robust results. Our algorithms segment the scene in semantically meaningful classes and recover some of the architectural structures of the scene such as symmetries. The obtained results are often both more accurate and more useful for applications. I will also briefly present some results on interactive immersive navigation of a video collection of an event.



Marc Pollefeys is a full professor and head of the Institute for Visual Computing of the Dept. of Computer Science of ETH Zurich which he joined in 2007. He currently also remains associated with the Dept. of Computer Science of the University of North Carolina at Chapel Hill where he started as an assistant professor in 2002 and became an associate professor in 2005. Dr. Pollefeys has received several prizes for his research, including a Marr prize, an NSF CAREER award, a Packard Fellowship and a ERC Starting Grant. Prof. Pollefeys is on the Editorial Board of the IEEE Transactions on Pattern Analysis and Machine Intelligence, the International Journal of Computer Vision and Foundations and Trends in Computer Graphics and Computer Vision.

Keynote Speaker 4 - Jong-Mo Seo (Wed. June 12. 09:00 – 10:00)
Title: Dynamic vs. Static vision and stereopsis
Abstract:
  Retina converts light to neural signal and performs basic image processing for the efficient handling of the visual information. This includes the color and brightness perception, edge detection and motion detection. Receptive field is the most well-known model for the explanation of these processes, and the retinal ganglion cell will generate the compressed visual information in the form of the frequency modulation. Static visual acuity is determined by the edge detection, and the dynamic visual acuity by the edge and the motion detection. Dynamic vision is synchronized with the vestibulocochlear information, and the mismatch between the two will make motion sickness. Stereopsis is determined by the monocular clue and binocular disparity, and the lateral geniculate nucleus plays an important role in the recognition of binocular disparity. In this talk, brief introduction for these mechanisms will be presented, and the related studies such as amblyopia, rivalry, dominance, after image and entopic phenomenon will be also discussed.



Jong-Mo Seo is an associate professor at the department of electrical engineering and computer sciences in Seoul National University (SNU) and an adjunct professor at the department of ophthalmology in Seoul National University Hospital (SNUH). He received the B.S., M.S., and Ph.D. degrees from the College of Medicine, SNU, and did his internship, residency, and retina fellowship at the department of ophthalmology in SNUH. He worked as a program planner for Order Communication System (OCS) of SNUH, and as a moderator of medical device interface team for Electronic Medical Record (EMR) of SNUH. He has been serving as a surgeon for animal experiment and also a moderator of the artificial retina project of South Korea since its establishment, in the year 2000. His research interests are the neural prosthesis, vision mechanism, and the computer-assisted diagnosis.

SPONSORS

DONORS

SCIENTIFIC SUPPORTER