Abstract The Gaussian mixture model(GMM) can use multiple Gaussian components to capture the variation information of image sets, and therefore it is a fine description method for image sets. Combining image set component symmetric positive definite descriptor]a component symmetric positive definite(SPD]model based on Gaussian mixture model(G-CSPD) is proposed. The image set is divided into sub-image sets with the same size, and the Gaussian mixture model of each sub-image set is calculated. A G-CSPD matrix in the form of the kernel matrix for all the sub-image sets is obtained, and the element in the matrix is used to denote the similarity between sub-image sets. The experimental results of 4 classification algorithms on 3 image sets show that G-CSPD is a more discriminative representation method for image sets.
[1]HARAND M, SALZMANN M, HARTLEY R. Dimensionality Reduction on SPD Manifolds: The Emergence of Geometry-Aware Methods. IEEE Transactions on Pattern Analysis and Machine Inte-lligence, 2018, 40(1): 48-62. [2]HUANG Z W, WANG R P, SHAN S G, et al. Projection Metric Learning on Grassmann Manifold with Application to Video Based Face Recognition // Proc of the IEEE Conference on Computer Vi-sion and Pattern Recognition. Washington, USA: IEEE, 2015: 140-149. [3]王锐,吴小俊.基于Log-Gabor滤波特征的黎曼流形图像集分类算法.模式识别与人工智能, 2017, 30(4): 377-384. (WANG R, WU X J. Riemannian Manifold Image Set Classification Algorithm Based on Log-Gabor Wavelet Features. Pattern Recognition and Artificial Intelligence, 2017, 30(4): 377-384.) [4]ZHU P F, ZHANG L, ZUO W M, et al.From Point to Set: Extend the Learning of Distance Metrics // Proc of the IEEE International Conference on Computer Vision. Washington, USA: IEEE, 2013: 2664-2671. [5]WANG W, WANG R P, HUANG Z W, et al.Discriminant Analysis on Riemannian Manifold of Gaussian Distributions for Face Reco-gnition with Image Sets // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2015: 2048-2057. [6]HARANDI M, SALZMANN M, BAKTASHMOTLAGH M. Beyond Gauss: Image-Set Matching on the Riemannian Manifold of PDFs // Proc of the IEEE International Conference on Computer Vision. Washington, USA: IEEE, 2015: 4112-4120. [7]WANG R P, GUO H M, DAVIS L S, et al.Covariance Discriminative Learning: A Natural and Efficient Approach to Image Set Classi-fication // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2012: 2496-2503. [8]HUANG Z W, WANG R P, SHAN S G, et al. Log-Euclidean Me-tric Learning on Symmetric Positive Definite Manifold with Application to Image Set Classification // Proc of the International Confe-rence on Machine Learning. New York, USA: ACM, 2015: 720-729. [9]HAMM J, LEE D D. Grassmann Discriminant Analysis: A Unifying View on Subspace-Based Learning // Proc of the 25th International Conference on Machine Learning. New York, USA: ACM, 2008: 376-383. [10]HARANDI M T, SANDERSON C, SHIRAZI S, et al.]Graph Embedding Discriminant Analysis on Grassmannian Manifolds for Improved Image Set Matching // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2011: 2705-2712. [11]HUANG Z W, VAN GOOL L. A Riemannian Network for SPD Matrix Learning[J/OL]. [2018-06-23]. https://arxiv.org/pdf/1608.04233.pdf. [12]HUANG Z W, WU J Q, VAN GOOL L. Building Deep Networks on Grassmann Manifolds[J/OL]. [2018-06-23]. https://arxiv.org/pdf/1611.05742.pdf. [13]CHEN Z Q, JIANG B, TANG J, et al. Image Set Representation and Classification with Attributed Covariate-Relation Graph Model and Graph Sparse Representation Classification. Neurocomputing, 2017, 226: 262-268. [14]HARANDI M T, SALZMANN M, HARTLEY R. From Manifold to Manifold: Geometry-Aware Dimensionality Reduction for SPD Matrices // Proc of the European Conference on Computer Vision. Berlin, Germany: Springer, 2014: 17-32. [15]CHEN K X, WU X J. Component SPD Matrices: A Low-Dimensional Discriminative Data Descriptor for Image Set Classification. Computational Visual Media, 2018, 4(3): 245-252. [16]DEMPSTER A P, LAIRD N M, RUBIN D B. Maximum Likelihood from Incomplete Data via the EM Algorithm. Journal of the Royal Statistical Society(Methodological), 1977, 39(1): 1-38. [17]LOVRIć M, MIN-OO M, RUH E A. Multivariate Normal Distributions Parametrized as a Riemannian Symmetric Space. Journal of Multivariate Analysis, 2000, 74(1): 36-48. [18]LI P H, WANG Q L, HUI Z, et al. Local Log-Euclidean Multivariate Gaussian Descriptor and Its Application to Image Classification. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(4): 803-817. [19]WANG R P, CHEN X L. Manifold Discriminant Analysis // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2009: 429-436. [20]ARSIGNY V, FILLARD P, PENNEC X, et al.Log-Euclidean Metrics for Fast and Simple Calculus on Diffusion Tensors. Magnetic Resonance in Medicine, 2006, 56(2): 411-421. [21]ARSIGNY V, FILLARD P, PENNEC X, et al. Geometric Means in a Novel Vector Space Structure on Symmetric Positive-Definite Matrices. SIAM Journal on Matrix Analysis and Applications, 2007, 29(1): 328-347. [22]DAI S L, MAN H. A Statistic Manifold Kernel with Graph Embe-dding Discriminant Analysis for Action and Expression Recognition // Proc of the IEEE International Conference on Image Processing. Washington, USA: IEEE, 2017: 1792-1796. [23]陈凯旋,吴小俊.面向图像集分类的切空间稀疏表示算法.中国图象图形学报, 2018, 23(7): 961-972. (CHEN K X, WU X J. Sparse Representation in Tangent Space for Image Set Classification. Journal of Image and Graphics, 2018, 23(7): 961-972.) [24]LI P H, WANG Q L, ZUO W M, et al. Log-Euclidean Kernels for Sparse Representation and Dictionary Learning // Proc of the IEEE International Conference on Computer Vision. Washington, USA: IEEE, 2013: 1601-1608. [25]FARAKI M, HARANDI M T, PORIKLI F. Image Set Classification by Symmetric Positive Semi-definite Matrices // Proc of the IEEE Winter Conference on Applications of Computer Vision. Washington, USA: IEEE, 2016. DOI: 10.1109/WACV.2016.7477621.
2018, Vol. 31 
