[BIC-announce] Yuri Boykov and Olga Veksler seminars
Kaleem Siddiqi
siddiqi at cim.mcgill.ca
Mon May 15 07:56:39 EDT 2017
Dear All,
Yuri Boykov and Olga Veksler of graph cuts fame will be visiting this Friday and will be giving back to back talks.
This is a terrific opportunity for students and colleagues alike…
Best,
Kaleem
---------------------------------------------------
Friday May 19th, McConnell Engineering 437, 11:00 am to 12:30
(2 back to back talks for 45 minutes) each.
Olga Veksler, Western University
Title: Adaptive and Move Making Auxiliary Cuts for Binary Pairwise Energies
Abstract:
Many computer vision problems require optimization of binary non-submodular energies. In this context, local iterative submodularization techniques based on trust region (LSA-TR) and auxiliary functions (LSA-AUX) have been recently proposed.
They achieve state-of-the-art-results on a number of computer vision applications. We extend the LSA-AUX framework in two directions. First, unlike LSA-AUX, which selects auxiliary functions based solely on the current solution, we propose to
incorporate several additional criteria. This results in tighter bounds for configurations that are more likely or closer to the current solution. Second, we propose move-making extensions of LSA-AUX which achieve tighter bounds by restricting
the search space. Finally, we evaluate our methods on several applications. We show that for each application at least one of our extensions significantly outperforms the original LSA-AUX. Moreover, the best extension of LSA-AUX is
comparable to or better than LSA-TR on four out of six applications.
——————————
Yuri Boykov, Western University
Title: Kernel Clustering meets Markov Random Fields
Abstract: The talk starts with an overview of standard kernel clustering techniques and their limitations. I particular we prove "Breiman's bias" under certain conditions and discuss its solutions. The talk also presents a new segmentation model combining common regularization energies, e.g. Markov Random Field (MRF) potentials, with standard pairwise clustering criteria like Normalized Cut (NC), average association (AA), etc. These clustering and regularization models are widely used in machine learning and computer vision, but they were not combined before due to significant differences in the corresponding optimization, e.g. spectral relaxation and combinatorial max-flow techniques. On the one hand, we show that common applications using MRF segmentation energies can benefit from a high-order NC term, e.g. enforcing balanced clustering of arbitrary high-dimensional image features combining color, texture, location, depth, motion, etc. On the other hand, standard clustering applications can benefit from an inclusion of common pairwise or higher-order MRF constraints, e.g. edge alignment, bin-consistency, label cost, etc. To address joint energies like NC+MRF, we propose efficient Kernel Cut algorithms based on bound optimization.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.bic.mni.mcgill.ca/pipermail/bic-announce/attachments/20170515/0019a20e/attachment.html>
More information about the BIC-announce
mailing list