选取论文的原则:
(1)会议论文,主要来源于以下会议:CVPR, ICCV, ECCV, BMVC, FG, ICIP, ICPR, WACV, ICASSP, MM, IJCAI, UAI, AAAI…(其实后面的几个会议都是打酱油的,AI类的期刊也木有高引CV文章,事实也是如此,没有收录ICML和NIPS等,见谅)
(2)发表在2000年以后,数据来源于微软学术搜索(http://academic.research.microsoft.com/),2012年12月初的检索结果
(3)检索次数大于200,PER>50,微软学术搜索的被检索次数未必精确,但能反应大概趋势,PER指意味着每年检索量,自发表次年算到2012年,PER=Cited/(2012-YEAR)
榜单Top 30如下,欢迎拾遗补缺:
[1] Rapid Object Detection using a Boosted Cascade of Simple Features (Citations: 3296, PER=299.64)
Paul A. Viola, Michael J. Jones @CVPR , vol. 1, pp. 511-518, 2001
[2] Histograms of Oriented Gradients for Human Detection (Citations: 1704, PER=243.43)
Navneet Dalal, Bill Triggs @CVPR , vol. 1, pp. 886-893, 2005
[3] SURF: Speeded-Up Robust Features (Citations: 1054, PER=175.67)
Herbert Bay, Tinne Tuytelaars, Luc J. Van Gool @ECCV , pp. 404-417, 2006
[4] Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories (Citations: 873, PER=145.5)
Svetlana Lazebnik, Cordelia Schmid, Jean Ponce @CVPR , vol. 2, pp. 2169-2178, 2006
[5] Object Class Recognition by Unsupervised Scale-Invariant Learning (Citations: 1071, PER=119)
Robert Fergus, Pietro Perona, Andrew Zisserman @CVPR , vol. 2, pp. 264-271, 2003
[6] Robust Real-Time Face Detection (Citations: 1092, PER=99.27)
Paul A. Viola, Michael J. Jones @ ICCV , 2001
[7] A Bayesian hierarchical model for learning natural scene categories (Citations: 677, PER=96.71)
Fei-Fei Li, Pietro Perona @CVPR , vol. 2, pp. 524-531, 2005
[8] Scalable Recognition with a Vocabulary Tree (Citations: 570, PER=95)
David Nistér, Henrik Stewénius @CVPR , vol. 2, pp. 2161-2168, 2006
[9] Real-Time Tracking of Non-Rigid Objects Using Mean Shift (Citations: 1132, PER=94.33)
Dorin Comaniciu, Visvanathan Ramesh, Peter Meer @CVPR , vol. 2, pp. 2142-149 vol.2, 2000
[10]Visual Categorization with Bags of Keypoints (Citations: 745, PER=93.13)
Gabriella Csurka, Christopher R. Dance, Lixin Fan, etc @ECCV , 2004
[11]Video Google: A Text Retrieval Approach to Object Matching in Videos (Citations: 790, PER=87.78)
Josef Sivic, Andrew Zisserman @ ICCV , pp. 1470-1477, 2003
[12]What Energy Functions Can Be Minimized via Graph Cuts? (Citations: 842, PER=84.2)
Vladimir Kolmogorov, Ramin Zabih @ECCV , pp. 65-81, 2002
[13]Overview of the Face Recognition Grand Challenge (Citations: 578, PER=82.57)
P. Jonathon Phillips, Patrick J. Flynn, W. Todd Scruggs, etc @CVPR , vol. 1, pp. 947-954, 2005
[14]Robust wide baseline stereo from maximally stable extremal regions (Citations: 810, PER=81)
Jiri Matas, Ondrej Chum, Martin Urban, etc @BMVC , vol. 1, 2002
[15]PCA-SIFT: A More Distinctive Representation for Local Image Descriptors (Citations: 639, PER=79.88)
Yan Ke, Rahul Sukthankar @CVPR , vol. 2, pp. 506-513, 2004
[16]Interactive Graph Cuts for Optimal Boundary and Region Segmentation of Objects in ND Images (Citations: 796, PER=72.36)
Yuri Y. Boykov, Marie-pierre Jolly @ ICCV , pp. 105-112, 2001
[17]An extended set of Haar-like features for rapid object detection (Citations: 710, PER=71)
Rainer Lienhart, Jochen Maydt @ICIP , vol. 1, pp. 900-903, 2002
[18]A Database of Human Segmented Natural Images and its Application to Evaluating Segmentation Algorithms and Measuring Ecological Statistics(Citations: 750, PER=68.18)
David R. Martin, Charless Fowlkes, Doron Tal, etc @ ICCV , pp. 416-425, 2001
[19]Detecting Pedestrians Using Patterns of Motion and Appearance (Citations: 584, PER=64.89)
Paul A. Viola, Michael J. Jones, Daniel Snow @ ICCV , pp. 734-741, 2003
[20]Object Recognition as Machine Translation: Learning a Lexicon for a Fixed Image Vocabulary (Citations: 603, PER=60.3)
Pinar Duygulu, Kobus Barnard, João F. G. De Freitas, etc @ECCV , pp. 97-112, 2002
[21]Real-Time Simultaneous Localisation and Mapping with a Single Camera (Citations: 527, PER=58.56)
Andrew J. Davison @ ICCV , pp. 1403-1410, 2003
[22]Recognizing Human Actions: A Local SVM Approach (Citations: 440, PER=55)
Christian Schüldt, Ivan Laptev, Barbara Caputo @ICPR , pp. 32-36, 2004
[23]Actions as Space-Time Shapes (Citations: 379, PER=54.14)
Moshe Blank, Lena Gorelick, Eli Shechtman, etc @ ICCV , vol. 2, pp. 1395-1402, 2005
[24]A Discriminatively Trained, Multiscale, Deformable Part Model (Citations: 215, PER=53.75)
Pedro F. Felzenszwalb, David A. Mcallester, Deva Ramanan @CVPR , pp. 1-8, 2008
[25]Non-parametric Model for Background Subtraction (Citations: 642, PER=53.5)
Ahmed M. Elgammal, David Harwood, Larry S. Davis @ECCV , pp. 751-767, 2000
[26]A Comparison and Evaluation of Multi-View Stereo Reconstruction Algorithms (Citations: 318, PER=53)
Steven M. Seitz, Brian Curless, James Diebel, etc @CVPR , vol. 1, pp. 519-528, 2006
[27]Comprehensive Database for Facial Expression Analysis (Citations: 636, PER=53)
Takeo Kanade, Yingli Tian, Jeffrey F. Cohn @FG , pp. 46-53, 2000
[28]Learning Realistic Human Actions from Movies (Citations: 211, PER=52.75)
Ivan Laptev, Marcin Marszalek, Cordelia Schmid, etc @CVPR , pp. 1-8, 2008
[29]Object Retrieval with Large Vocabularies and Fast Spatial Matching (Citations: 258, PER=51.6)
James Philbin, Ondrej Chum, Michael Isard, etc @CVPR , 2007
[30]Statistical Shape Influence in Geodesic Active Contours (Citations: 616, PER=51.33)
Michael E. Leventon, W. Eric L. Grimson, Olivier D. Faugeras @CVPR , vol. 1, pp. 1316-1323, 2000
简单小结:
总共30篇会议文章,其中CVPR 14篇,ECCV 5篇,ICCV 7篇,BMVC、FG、ICIP、ICPR各1篇。每年发表的论文数如图所示。
我个人比较关注3个方面,(1)检测Detection(比如基于Harr-like的Viola-Jones的3篇文章,[1][6][19],还有[17],基于HOG的[2],以及在PASCAL VOC上大放光芒的DPM[24]);(2)基于Bag of Words (Features)和LDA的图像检索Retrieval和类别学习Categorization(比如基于BoW的[10][11],之后的[8][29],以及非常流行的SPM[4],基于LDA的[7]等);(3)特征Features,比如说提到的Harr-like,SURF[3],HOG[2],MSER[14],PCA-SIFT[15]等。
值得注意的是,排在前面的大部分作者,都是CV界的成名人物,非常值得关注他们的其他一些论文。
PS:[12]的作者Ramin Zabih从TPAMI的主编位置上退下,[20]的作者David Forsyth成为TPAMI新的掌舵者,大家有什么看法?
最后额外推荐几篇ML的方法,来自于ICML和NIPS。
[1] Reinforcement learning: An introduction (Citations: 5592)
Richard S. Sutton, Andrew G. Barto @NIPS , vol. 9, no. 5, pp. 1054-1054, 1998
[2] Experiments with a New Boosting Algorithm (Citations: 2489)
Yoav Freund, Robert E. Schapire @ICML , pp. 148-156, 1996
[3] Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data (Citations: 2436)
John D. Lafferty, Andrew McCallum, Fernando C. N. Pereira @ICML , pp. 282-289, 2001
[4] Latent Dirichlet Allocation (Citations: 1548)
David M. Blei, Andrew Y. Ng, Michael I. Jordan @NIPS , pp. 601-608, 2001
[5] On Spectral Clustering: Analysis and an algorithm (Citations: 1230)
Andrew Y. Ng, Michael I. Jordan, Yair Weiss @NIPS , pp. 849-856, 2001
[6] Algorithms for Nonnegative Matrix Factorization (Citations: 1162)
Daniel D. Lee, H. Sebastian Seung @NIPS , vol. 13, pp. 556-562, 2000
[7] Semi-Supervised Learning Using Gaussian Fields and Harmonic Functions (Citations: 665)
Xiaojin Zhu, Zoubin Ghahramani, John D. Lafferty @ICML , pp. 912-919, 2003
2 最有影响力的20篇计算机视觉期刊论文
选取论文的原则:
(1)期刊论文,主要来源于以下期刊:TPAMI,IJCV,TIP,CVIU,IVC,MVA,PR,JMIV,IJPRAI…
(2)发表在2000年以后
(3)SCI检索次数大于1000,来源于Web of Science数据库,2012年12月初的检索结果
Top 20 榜单原文请见http://www.cnblogs.com/youth0826/archive/2012/12/04/2801481.html