动态多视角稀疏表示2012.pdfVIP

Pattern Recognition 45 (2012) 1290–1298 Contents lists available at SciVerse ScienceDirect Pattern Recognition journal homepage: /locate/pr Joint dynamic sparse representation for multi-view face recognition Haichao Zhang a,b,n, Nasser M. Nasrabadi c, Yanning Zhang a, Thomas S. Huang b a School of Computer Science, Northwestern Polytechnical University, Xi’an 710129, China b Beckman Institute, University of Illinois at Urbana-Champaign, IL, USA c U.S. Army Research Laboratory, 2800 Powder Mill Road, Adelphi, MD, USA a r t i c l e i n f o a b s t r a c t Article history: We consider the problem of automatically recognizing a human face from its multi-view images with Received 27 June 2011 unconstrained poses. We formulate the multi-view face recognition task as a joint sparse representa- Received in revised form tion model and take advantage of the correlations among the multiple views for face recognition using 7 September 2011 a novel joint dynamic sparsity prior. The proposed joint dynamic sparsity prior promotes shared joint Accepted 9 September 2011 sparsity patterns among the multiple sparse representation vectors at class-level, while allowing Available online 22 September 2011 distinct sparsity patterns at atom-level within each class to facilitate a flexible representation. Extensive Keywords: experiments on the CMU Multi-PIE face database are conducted to verify th