正在加载图片...
·146· 智能系统学报 第11卷 sampling[J].IEEE,signal processing magazine,2008,25 4 Conclusions (2):21-30. In this study,to deal with sparsity and instability [7]CANDES E J,ROMBERG J.TAO J.Robust uncertainty in the optimization problem and the high time principles:exact signal reconstruction from highly incom- complexity of the NLSSSC-tracker [11],we propose a plete frequency information[].IEEE transactions on infor- novel efficient tracker,i.e.,the Euclidean local-struc- mation theory,2006,52(2):489-509. [8]MEI Xue,LING Haibin,WU Yi,et al.Minimum error ture constraint based sparse coding (ELSSC).Our new bounded efficient /1 tracker with occlusion detection[Cl// algorithm is a l-tracker with a reconstructed over-com- Proceedings of IEEE Conference on Computer Vision and plete dictionary,which is different from that in the o- Pattern Recognition (CVPR).Colorado,USA,2011:1257- riginal I,-tracker and NLSSSC-tracker.Moreover,we 1264. simplify the large-scale I-optimization problem in our [9]MEI Xue,LING Haibin.Robust visual tracking and vehicle tracker to a much smaller one in our improved ELSSC- classification via sparse representation[].IEEE transac- tracker. tions on pattern analysis and machine intelligence,2011,33 Compared with the original 1,-tracker,our (11):2259-2272. [10 XU Huan,CARAMANIS C,MANNOR S.Sparse algo- ELSSC-tracker introduces the structure information a- rithms are not stable:a no-free-lunch theorem].IEEE mong the candidate regions generated by the Bayesian transactions on pattern analysis and machine intelligence, inference to the I-tracker,similar to that in the 2011,34(1):187-193. NLSSSC-tracker.With our derivation,the optimization [11]LU Xiaoqiang,YUAN Yuan,LU Pingkun,et al.Robust procedure of our tracker (Eg.(10))can be solved as visual tracking with discriminative sparse learning[J].Pat- that in the I-optimization but very differently from that tern recognition,2013,46(7):1762-1771. in the NLSSSC.Furthermore,our improved tracker is [12]YANG Jian,ZHANG Lei,XU Yong,et al.Beyond sparsi- much more efficient than the I-tracker and NLSSSC- ty:the role of L-optimizer in pattern classification [J]. tracker.Our experiments demonstrate the sparsity,sta- Pattern recognition,2012,45(3):1104-1118. [13]DAUBECHIES I,DEFRISE M,DE MOL C.An iterative bility,and efficiency of our tracker. thresholding algorithm for linear inverse problems with a References sparsity constraint[J].Communications on pure and ap- plied mathematics,2004,57(11):1413-1457. [1]ZHANG Shengping,YAO Hongxun,SUN Xin,et al.Sparse [14]ROSS D A,LIM J,LIN R S,et al.Incremental learning coding based visual tracking:review and experimental com- for robust visual tracking[J.International journal of com- parison[]].Pattern recognition,2013,46(7):1772-1788. puter vision,2008,77(1-3):125-141. [2]YILMAZ A,JAVED O,SHAH M.Object tracking:a sur- [15]PORIKLI F,TUZEL O,MEER P.Covariance tracking u- vey[]].ACM computing surveys (CSUR),2006,38(4): sing model update based on lie algebra[C]//Proceedings 1-.45. of IEEE Computer Society Conference on Computer Vision [3]KALAL Z,MIKOLAJCZYK K,MATAS J.Tracking-learn- and Pattern Recognition.New York,USA,2006:728- ing-detection[].IEEE transactions on pattern analysis and 735. machine intelligence,2012,34(7):1409-1422. [16]EVERINGHAM M,VAN GOOL L,WILLIAMS C K I,et [4]AVIDAN S.Ensemble tracking[J].IEEE transactions on al.The pascal visual object classes (VOC)challenge[J]. pattern analysis and machine intelligence,2007,29(2): International journal of computer vision,2010,88(2): 261-271. 303.338. [5]BABENKO B,YANG M H,BELONGIE S.Visual tracking [17]WU Yi,LIM J,YANG M H.Online object tracking:A with online multiple instance learning[C]//Proceedings of benchmark[C]//Proceedings of IEEE Conference on Com- IEEE Conference on Computer Vision and Pattern Recogni- puter Vision and Pattern Recognition (CVPR).Portland, tion (CVPR).Miami,USA.2009:983-990. USA,2013:2411-2418 [6]CANDES EJ,WAKIN M B.An introduction to compressive [18]KRISTAN M,PAUGFELDER R,LEONARDIS A,et al.4 Conclusions In this study, to deal with sparsity and instability in the l 1 ⁃optimization problem [10⁃12] and the high time complexity of the NLSSSC⁃tracker [11], we propose a novel efficient tracker, i.e., the Euclidean local⁃struc⁃ ture constraint based sparse coding (ELSSC). Our new algorithm is a l 1 ⁃tracker with a reconstructed over⁃com⁃ plete dictionary, which is different from that in the o⁃ riginal l 1 ⁃tracker and NLSSSC⁃tracker. Moreover, we simplify the large⁃scale l 1 ⁃optimization problem in our tracker to a much smaller one in our improved ELSSC⁃ tracker. Compared with the original l 1 ⁃tracker, our ELSSC⁃tracker introduces the structure information a⁃ mong the candidate regions generated by the Bayesian inference to the l 1 ⁃tracker, similar to that in the NLSSSC⁃tracker. With our derivation, the optimization procedure of our tracker (Eq.(10)) can be solved as that in the l 1 ⁃optimization but very differently from that in the NLSSSC. Furthermore, our improved tracker is much more efficient than the l 1 ⁃tracker and NLSSSC⁃ tracker. Our experiments demonstrate the sparsity, sta⁃ bility, and efficiency of our tracker. References [1]ZHANG Shengping, YAO Hongxun, SUN Xin, et al. Sparse coding based visual tracking: review and experimental com⁃ parison[J]. Pattern recognition, 2013, 46(7): 1772⁃1788. [2]YILMAZ A, JAVED O, SHAH M. Object tracking: a sur⁃ vey[J]. ACM computing surveys (CSUR), 2006, 38(4): 1⁃45. [3]KALAL Z, MIKOLAJCZYK K, MATAS J. Tracking⁃learn⁃ ing⁃detection[J]. IEEE transactions on pattern analysis and machine intelligence, 2012, 34(7): 1409⁃1422. [4] AVIDAN S. Ensemble tracking [ J]. IEEE transactions on pattern analysis and machine intelligence, 2007, 29( 2): 261⁃271. [5]BABENKO B, YANG M H, BELONGIE S. Visual tracking with online multiple instance learning[C] / / Proceedings of IEEE Conference on Computer Vision and Pattern Recogni⁃ tion (CVPR). Miami, USA, 2009: 983⁃990. [6]CANDÈS E J, WAKIN M B. An introduction to compressive sampling[J]. IEEE, signal processing magazine, 2008, 25 (2): 21⁃30. [7] CANDÈS E J, ROMBERG J, TAO J. Robust uncertainty principles: exact signal reconstruction from highly incom⁃ plete frequency information[J]. IEEE transactions on infor⁃ mation theory, 2006, 52(2): 489⁃509. [8] MEI Xue, LING Haibin, WU Yi, et al. Minimum error bounded efficient l1 tracker with occlusion detection[C] / / Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Colorado, USA, 2011:1257⁃ 1264. [9]MEI Xue, LING Haibin. Robust visual tracking and vehicle classification via sparse representation [ J]. IEEE transac⁃ tions on pattern analysis and machine intelligence, 2011, 33 (11): 2259⁃2272. [10] XU Huan, CARAMANIS C, MANNOR S. Sparse algo⁃ rithms are not stable: a no⁃free⁃lunch theorem[ J]. IEEE transactions on pattern analysis and machine intelligence, 2011, 34(1): 187⁃193. [11]LU Xiaoqiang, YUAN Yuan, LU Pingkun, et al. Robust visual tracking with discriminative sparse learning[J]. Pat⁃ tern recognition, 2013, 46(7): 1762⁃1771. [12]YANG Jian, ZHANG Lei, XU Yong, et al. Beyond sparsi⁃ ty: the role of L1 ⁃optimizer in pattern classification [ J]. Pattern recognition, 2012, 45(3): 1104⁃1118. [13]DAUBECHIES I, DEFRISE M, DE MOL C. An iterative thresholding algorithm for linear inverse problems with a sparsity constraint[ J]. Communications on pure and ap⁃ plied mathematics, 2004, 57(11): 1413⁃1457. [14]ROSS D A, LIM J, LIN R S, et al. Incremental learning for robust visual tracking[J]. International journal of com⁃ puter vision, 2008, 77(1⁃3): 125⁃141. [15]PORIKLI F, TUZEL O, MEER P. Covariance tracking u⁃ sing model update based on lie algebra[C] / / Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition. New York, USA, 2006: 728⁃ 735. [16]EVERINGHAM M, VAN GOOL L, WILLIAMS C K I, et al. The pascal visual object classes (VOC) challenge[ J]. International journal of computer vision, 2010, 88 ( 2): 303⁃338. [17]WU Yi, LIM J, YANG M H. Online object tracking: A benchmark[C] / / Proceedings of IEEE Conference on Com⁃ puter Vision and Pattern Recognition (CVPR). Portland, USA, 2013: 2411⁃2418. [18] KRISTAN M, PflUGFELDER R, LEONARDIS A, et al. ·146· 智 能 系 统 学 报 第 11 卷
<<向上翻页向下翻页>>
©2008-现在 cucdc.com 高等教育资讯网 版权所有