Stochastic Covariance Compression

Covariance matrices are an effective way to capture global spread across local interest points in images. Often, these image descriptors are more compact, robust and informative than, for example, bags of visual words. However, they are symmetric and positive definite (SPD) and therefore live on a non-Euclidean Riemannian manifold, which gives rise to non-Euclidean metrics. These are slow to compute and can make the use of covariance features prohibitive in many settings, in particular k-nearest neighbors (kNN) classification. In this paper we present Stochastic Covariance Compression, an algorithm that compresses a data set of SPD matrices to a much smaller set with similar kNN characteristics. We show that we can reduce the data sets to 1/6 and in some cases even up to 1/50 of their original size, while approximately matching the test error of full kNN classification. In fact, because the compressed set is learned to perform well on kNN tasks, it sometimes even outperforms the original data set, while requiring only a fraction of the space and drastically reduced test-time computation.

[1]  Fabrizio Angiulli,et al.  Fast condensed nearest neighbor rule , 2005, ICML.

[2]  Godfried T. Toussaint,et al.  Proximity Graphs for Nearest Neighbor Decision Rules: Recent Progress , 2002 .

[3]  Dieter Fox,et al.  A large-scale hierarchical multi-view RGB-D object dataset , 2011, 2011 IEEE International Conference on Robotics and Automation.

[4]  Teuvo Kohonen,et al.  Improved versions of learning vector quantization , 1990, 1990 IJCNN International Joint Conference on Neural Networks.

[5]  Brian C. Lovell,et al.  Sparse Coding and Dictionary Learning for Symmetric Positive Definite Matrices: A Kernel Approach , 2012, ECCV.

[6]  John F. Canny,et al.  A Computational Approach to Edge Detection , 1986, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[7]  Hongdong Li,et al.  Kernel Methods on the Riemannian Manifold of Symmetric Positive Definite Matrices , 2013, 2013 IEEE Conference on Computer Vision and Pattern Recognition.

[8]  Simon Kasif,et al.  Best-Case Results for Nearest-Neighbor Learning , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[9]  C. G. Hilborn,et al.  The Condensed Nearest Neighbor Rule , 1967 .

[10]  Sergio Bermejo,et al.  Adaptive soft k-nearest-neighbor classifiers , 1999, Pattern Recognit..

[11]  Kenneth I. Laws,et al.  Rapid Texture Identification , 1980, Optics & Photonics.

[12]  Rich Caruana,et al.  Model compression , 2006, KDD '06.

[13]  I. Tomek,et al.  Two Modifications of CNN , 1976 .

[14]  Xuelong Li,et al.  Gabor-Based Region Covariance Matrices for Face Recognition , 2008, IEEE Transactions on Circuits and Systems for Video Technology.

[15]  Geoffrey E. Hinton,et al.  Stochastic Neighbor Embedding , 2002, NIPS.

[16]  Geoffrey E. Hinton,et al.  Neighbourhood Components Analysis , 2004, NIPS.

[17]  Fatih Murat Porikli,et al.  Covariance Tracking using Model Update Based on Lie Algebra , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[18]  Fatih Murat Porikli,et al.  Region Covariance: A Fast Descriptor for Detection and Classification , 2006, ECCV.

[19]  Geoffrey E. Hinton,et al.  Visualizing Data using t-SNE , 2008 .

[20]  Antonio Torralba,et al.  Modeling the Shape of the Scene: A Holistic Representation of the Spatial Envelope , 2001, International Journal of Computer Vision.

[21]  Suvrit Sra,et al.  A new metric on the manifold of kernel matrices with application to matrix geometric means , 2012, NIPS.

[22]  Krista A. Ehinger,et al.  SUN database: Large-scale scene recognition from abbey to zoo , 2010, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[23]  M. Narasimha Murty,et al.  An incremental prototype set building technique , 2002, Pattern Recognit..

[24]  René Vidal,et al.  Clustering and dimensionality reduction on Riemannian manifolds , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[25]  Cordelia Schmid,et al.  Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[26]  Christine Decaestecker,et al.  Finding prototypes for nearest neighbour classification by means of gradient descent and deterministic annealing , 1997, Pattern Recognit..

[27]  Anoop Cherian,et al.  Efficient similarity search for covariance matrices via the Jensen-Bregman LogDet Divergence , 2011, 2011 International Conference on Computer Vision.

[28]  G. Gates The Reduced Nearest Neighbor Rule , 1998 .

[29]  Xavier Pennec,et al.  A Riemannian Framework for Tensor Computing , 2005, International Journal of Computer Vision.

[30]  James M. Rehg,et al.  Learning sparse covariance patterns for natural scenes , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[31]  Anoop Cherian,et al.  Riemannian Sparse Coding for Positive Definite Matrices , 2014, ECCV.

[32]  David W. Jacobs,et al.  Riemannian Metric Learning for Symmetric Positive Definite Matrices , 2015, ArXiv.

[33]  TuzelOncel,et al.  Pedestrian Detection via Classification on Riemannian Manifolds , 2008 .

[34]  Fatih Murat Porikli,et al.  Pedestrian Detection via Classification on Riemannian Manifolds , 2008, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[35]  Kilian Q. Weinberger,et al.  Distance Metric Learning for Large Margin Nearest Neighbor Classification , 2005, NIPS.

[36]  Peter E. Hart,et al.  Nearest neighbor pattern classification , 1967, IEEE Trans. Inf. Theory.

[37]  Eli Shechtman,et al.  Matching Local Self-Similarities across Images and Videos , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[38]  Lawrence Cayton,et al.  Fast nearest neighbor retrieval for bregman divergences , 2008, ICML '08.

[39]  Stephen Tyree,et al.  Stochastic Neighbor Compression , 2014, ICML.

[40]  Don Coppersmith,et al.  Matrix multiplication via arithmetic progressions , 1987, STOC.