Combining Classifiers for Improved Multilabel Image Classification

We propose a stacking-like method for multilabel image classification. Our approach combines the output of binary base learners, which use different features for image description, in a simple and straightforward way: The confidence values of the base learners are fed into a support vector machine (SVM) in order to improve prediction accuracy. Experiments on the datasets of the Pascal Visual Object Classes challenges (VOC) of 2006 and 2007 show that our method significantly improves over the performance of the base learners. Our approach also works better than more naive approaches for combining features or classifiers.

[1]  Luc Van Gool,et al.  Affine/ Photometric Invariants for Planar Intensity Patterns , 1996, ECCV.

[2]  Hans-Peter Kriegel,et al.  Using Support Vector Machines for Classifying Large Sets of Multi-Represented Objects , 2004, SDM.

[3]  Florent Perronnin,et al.  Fisher Kernels on Visual Vocabularies for Image Categorization , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[4]  Sunita Sarawagi,et al.  Discriminative Methods for Multi-labeled Classification , 2004, PAKDD.

[5]  Ayhan Demiriz,et al.  Linear Programming Boosting via Column Generation , 2002, Machine Learning.

[6]  Tao Mei,et al.  Correlative multi-label video annotation , 2007, ACM Multimedia.

[7]  Remco C. Veltkamp,et al.  Spatial pyramids and two-layer stacking SVM classifiers for image categorization: A comparative study , 2009, 2009 International Joint Conference on Neural Networks.

[8]  David H. Wolpert,et al.  Stacked generalization , 1992, Neural Networks.

[9]  Ian H. Witten,et al.  Issues in Stacked Generalization , 2011, J. Artif. Intell. Res..

[10]  Thomas G. Dietterich.  Machine-Learning Research Four Current Directions , 1997 .

[11]  Ryan M. Rifkin,et al.  In Defense of One-Vs-All Classification , 2004, J. Mach. Learn. Res..

[12]  Jiri Matas,et al.  On Combining Classifiers , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[13]  Yoav Freund,et al.  Boosting a weak learning algorithm by majority , 1990, COLT '90.

[14]  Martin Antenreiter,et al.  Visual Classification of Images by Learning Geometric Appearances Through Boosting , 2006, ANNPR.

[15]  Peter Auer,et al.  Object recognition using segmentation for feature detection , 2004, ICPR 2004.

[16]  Robert E. Schapire,et al.  The strength of weak learnability , 1990, Mach. Learn..

[17]  John Shawe-Taylor,et al.  A Column Generation Algorithm For Boosting , 2000, ICML.

[18]  Bernard Zenko,et al.  Is Combining Classifiers with Stacking Better than Selecting the Best One? , 2004, Machine Learning.

[19]  Matthijs C. Dorst.  Distinctive Image Features from Scale-Invariant Keypoints , 2011 .

[20]  Yiannis Kompatsiaris,et al.  An Empirical Study of Multi-label Learning Methods for Video Annotation , 2009, 2009 Seventh International Workshop on Content-Based Multimedia Indexing.

[21]  C. Schmid,et al.  Indexing based on scale invariant interest points , 2001, Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001.

[22]  Jianguo Zhang,et al.  The PASCAL Visual Object Classes Challenge , 2006 .

[23]  Evon M. O. Abu-Taieh,et al.  Comparative Study , 2020, Definitions.

[24]  Robert C. Holte,et al.  Very Simple Classification Rules Perform Well on Most Commonly Used Datasets , 1993, Machine Learning.

[25]  Yan Ke,et al.  PCA-SIFT: a more distinctive representation for local image descriptors , 2004, CVPR 2004.

[26]  Nello Cristianini,et al.  An Introduction to Support Vector Machines and Other Kernel-based Learning Methods , 2000 .

[27]  Christopher J. Merz,et al.  Using Correspondence Analysis to Combine Classifiers , 1999, Machine Learning.