高阶马尔科夫随机场及其在场景理解中的应用

被引:23
|
作者
余淼 [1 ,2 ]
胡占义 [1 ]
机构
[1] 中国科学院自动化研究所模式识别国家重点实验室
[2] 中原工学院电子信息学院
基金
国家高技术研究发展计划(863计划);
关键词
高阶马尔科夫随机场; 能量模型; 能量优化; 参数学习; 场景理解;
D O I
10.16383/j.aas.2015.c140684
中图分类号
TP391.41 [];
学科分类号
080203 ;
摘要
与传统的一阶马尔科夫随机场(Markov random field,MRF)相比,高阶马尔科夫随机场能够表达更加复杂的定性和统计性先验信息,在模型的表达能力上具有更大的优势.但高阶马尔科夫随机场对应的能量函数优化问题更为复杂.同时其模型参数数目的爆炸式增长使得选择合适的模型参数也成为了一个非常困难的问题.近年来,学术界在高阶马尔科夫随机场的能量模型的建模、优化和参数学习三个方面进行了深入的探索,取得了很多有意义的成果.本文首先从这三个方面总结和介绍了目前在高阶马尔科夫随机场研究上取得的主要成果,然后介绍了高阶马尔科夫随机场在图像理解和三维场景理解中的应用现状.
引用
收藏
页码:1213 / 1234
页数:22
相关论文
共 20 条
  • [1] Inference Methods for CRFs with Co-occurrence Statistics
    Ladicky, L'ubor
    Russell, Chris
    Kohli, Pushmeet
    Torr, Philip H. S.
    [J]. INTERNATIONAL JOURNAL OF COMPUTER VISION, 2013, 103 (02) : 213 - 225
  • [2] Joint Optimization for Object Class Segmentation and Dense Stereo Reconstruction
    Ladicky, Lubor
    Sturgess, Paul
    Russell, Chris
    Sengupta, Sunando
    Bastanlar, Yalin
    Clocksin, William
    Torr, Philip H. S.
    [J]. INTERNATIONAL JOURNAL OF COMPUTER VISION, 2012, 100 (02) : 122 - 133
  • [3] Fast Approximate Energy Minimization with Label Costs
    Delong, Andrew
    Osokin, Anton
    Isack, Hossam N.
    Boykov, Yuri
    [J]. INTERNATIONAL JOURNAL OF COMPUTER VISION, 2012, 96 (01) : 1 - 27
  • [4] Harmony Potentials
    Boix, Xavier
    Gonfaus, Josep M.
    van de Weijer, Joost
    Bagdanov, Andrew D.
    Serrat, Joan
    Gonzalez, Jordi
    [J]. INTERNATIONAL JOURNAL OF COMPUTER VISION, 2012, 96 (01) : 83 - 102
  • [5] The Pascal Visual Object Classes (VOC) Challenge
    Everingham, Mark
    Van Gool, Luc
    Williams, Christopher K. I.
    Winn, John
    Zisserman, Andrew
    [J]. INTERNATIONAL JOURNAL OF COMPUTER VISION, 2010, 88 (02) : 303 - 338
  • [6] Robust Higher Order Potentials for Enforcing Label Consistency
    Kohli, Pushmeet
    Ladicky, L'ubor
    Torr, Philip H. S.
    [J]. INTERNATIONAL JOURNAL OF COMPUTER VISION, 2009, 82 (03) : 302 - 324
  • [7] Efficient belief propagation for higher-order cliques using linear constraint nodes
    Potetz, Brian
    Lee, Tai Sing
    [J]. COMPUTER VISION AND IMAGE UNDERSTANDING, 2008, 112 (01) : 39 - 54
  • [8] TextonBoost for Image Understanding: Multi-Class Object Recognition and Segmentation by Jointly Modeling Texture, Layout, and Context[J] . Jamie Shotton,John Winn,Carsten Rother,Antonio Criminisi.International Journal of Computer Vision . 2009 (1)
  • [9] Efficient belief propagation for early vision
    Felzenszwalb, Pedro F.
    Huttenlocher, Daniel P.
    [J]. INTERNATIONAL JOURNAL OF COMPUTER VISION, 2006, 70 (01) : 41 - 54
  • [10] Pseudo-Boolean optimization[J] . Endre Boros,Peter L. Hammer.Discrete Applied Mathematics . 2002 (1)