See All by Looking at A Few: Sparse Modeling for Finding Data Exemplars 

Given a set of data points, we consider the problem of finding a subset of the data points, called representatives or exemplars, that efficiently represent the entire dataset. In our formulation, each data point is expressed as a linear combination of the exemplars, which are selected from the data via convex optimization. In general, we do not assume that the data are low-rank or distributed around cluster centers. However, when the data do come from a collection of low-rank models, we show that our method automatically selects a few representatives from each low-rank model. Our framework can be extended to detect and reject outliers in datasets, and to efficiently deal with new observations. Our framework can also be extended to deal with pairwise dissimilarities, rather than data points. In this case, we show that 
when data points are distributed around multiple clusters according to the dissimilarities, the data points in each cluster select representatives only from that cluster. The proposed framework and theoretical foundations are illustrated with examples in text summarization, video summarization and image classification using exemplars. This is joint work with Elhsan Elhamifar and Guillermo Sapiro.

Prof. Rene Vidal received his B.S. in EE (valedictorian) from the Universidad Catolica de Chile in 1997, and his M.S. and Ph.D. in EECS from UC Berkeley in 2000 and 2003, respectively. In 2004, he joined the faculty of the Center for Imaging Science and the Department of Biomedical Engineering of Johns Hopkins University. His areas of research are biomedical image analysis, computer vision, machine learning, signal processing, hybrid systems, and robotics. Dr. Vidal is or has been Associate Editor of the IEEE Transactions on Pattern Analysis and Machine Intelligence, the SIAM Journal on Imaging Sciences and the Journal of Mathematical Imaging and Vision. He has also served as Program Chair for ICCV 2015, CVPR 2014, WMVC 2009 and PSIVT 2007, and Area Chair for MICCAI 2013 and 2014, ICCV 2007, 2011 and 2013, and CVPR 2005 and 2013. He has received many awards for his work including the 2012 J.K. Aggarwal Prize, the 2009 ONR Young Investigator Award, the 2009 Sloan Research Fellowship, the 2005 NFS CAREER Award, and best paper awards at ICCV-3DRR 2013, PSIVT 2013, CDC 2012, MICCAI 2012, CDC 2011 and ECCV 2004. He is a fellow of the IEEE and a member of the ACM and SIAM.