Uwe Franke, Daimler Research & Development
Computer Vision for Autonomous Cars
" Recent Mercedes Benz cars offer a powerful STEREO camera system that sets new standards in vehicle safety and comfort. Modern dense disparity estimation combined with top-performing pedestrian classification allow for fully autonomous emergency braking. The system works for a speed range up to 72km/h – both day and night. Even more, for the first time Mercedes’ Intelligent Drive offers autonomous driving at low speeds in traffic jams. Currently, the research is heading towards autonomous driving on highways as well as in cities.
The talk will sketch the state-of-the-art in robust Computer Vision for driver assistance. In addition, recent developments in Image Understanding that hopefully pave the way towards accident-free driving will be presented."
Uwe Franke received his Diploma degree and his PhD degree both in electrical communictions engineering from Aachen Technical University in 1983 and 1988. Since 1989 he is with Daimler Research & Development. He developed Daimler's lane departure warning system and has been working on stereo vision since 1996. Since 2000 he is head of Daimler's Image Understanding group. The algorithms developed by this group are the basis for Daimler's Stereo Camera based safety systems that have become commercially available in Mercedes vehicles.
Uwe Franke is a member of the technical committee of the German Association for Pattern Recognition and was program chair of the IEEE Intelligent Vehicles Symposium IV 2002 in Versailles (France). He was nominated for the German Future Prize in 2011 and was awarded the Karl Heinz Beckurts-Prize in 2012.
[Dr. Uwe Franke's web page]
Kenichi Kanatani, Okayama University
Overview of Optimization Techniques for Geometric Estimation
"I summarize techniques for optimal geometric estimation from noisy observations for computer vision applications. I start with estimation techniques based on minimization of given cost functions: least squares (LS), maximum likelihood (ML), which includes reprojection error minimization (Gold Standard) as a special case, and Sampson error minimization. I then formulate estimation techniques not based on minimization of any cost function: iterative reweight, renormalization, and hyper-renormalization. Showing numerical examples, I conclude that hyper-renormalization is robust to noise and currently is the best method."
Professor Kenichi Kanatani received his B.E., M.S., and Ph.D. in applied
mathematics from the University of Tokyo in 1972, 1974 and 1979, respectively. After serving as Professor of computer science at Gunma University, Gunma, Japan, till March 2001, he moved to Okayama University, where he was Professor of computer science till March 2013. He is now Research Professor of Okayama University.
He is the author of "Group-Theoretical Methods in Image Understanding" (Springer 1990), "Geometric Computation for Machine Vision" (Oxford University Press, 1993) and "Statistical Optimization for Geometric Computation: Theory and Practice" (Elsevier 1996; reprinted Dover 2005). He received the best paper awards from IPSJ (1987), IEICE (2005), and PSIVT 2009. He is an IEEE and IEICE Fellow.
[Prof. Kenichi Kanatani's web page]
Yasuyuki Matsushita, Microsoft Research Asia
Toward High-Fidelity 3D Reconstruction: A Photometric Approach
" Recent years have shown tremendous advances on 3D reconstruction in computer vision and sensor technologies. However, most of these techniques are limited to estimate a coarse-scale shape that lacks fine-details of the surface. In this talk, I will discuss a photometric 3D reconstruction approach with which fine-details are faithfully recovered using shading cues in the form of surface normal. Specifically, I will talk about recent photometric stereo techniques that robustly and accurately recover surface normal of a scene that has diverse reflectance properties."
Yasuyuki Matsushita received his B.S., M.S. and Ph.D. degrees in EECS from the University of Tokyo in 1998, 2000, and 2003, respectively. He joined Microsoft Research Asia in April 2003. He is a Lead Researcher in Visual Computing Group. His areas of research are Computer Vision (photometric techniques, such as radiometric calibration, photometric stereo, shape-from-shading), Computer Graphics (image relighting, video analysis and synthesis).
He is on the editorial board member of IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), International Journal of Computer Vision (IJCV), IPSJ Journal of Computer Vision and Applications (CVA), The Visual Computer Journal, and Encyclopedia of Computer Vision. He served/is serving as a Program Co-Chair of PSIVT 2010, 3DIMPVT 2011, and ACCV 2012. He is appointed as a Guest Associate Professor at Osaka University (April 2010-), visiting Associate Professor at National Institute of Informatics (April 2011-) and Tohoku University (April 2012 - ), Japan. He is a senior member of IEEE.
[Dr. Yasuyuki Matsushita's web page]