Bandwidth selection for kernel estimation in mixed multi-dimensional spaces

Kernel estimation techniques, such as mean shift, suffer from one major drawback: the kernel bandwidth selection. The bandwidth can be fixed for all the data set or can vary at each points. Automatic bandwidth selection becomes a real challenge in case of multidimensional heterogeneous features. This paper presents a solution to this problem. It is an extension of \cite{Comaniciu03a} which was based on the fundamental property of normal distributions regarding the bias of the normalized density gradient. The selection is done iteratively for each type of features, by looking for the stability of local bandwidth estimates across a predefined range of bandwidths. A pseudo balloon mean shift filtering and partitioning are introduced. The validity of the method is demonstrated in the context of color image segmentation based on a 5-dimensional space.

[1]  Yee Leung,et al.  Clustering by Scale-Space Filtering , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  Douglas G. Simpson.  Nearest-neighbor variance estimation (NNVE): Robust covariance estimation via nearest-neighbor cleaning - Comment , 2002 .

[3]  L. Breiman,et al.  Variable Kernel Estimates of Multivariate Densities , 1977 .

[4]  Dorin Comaniciu,et al.  The Variable Bandwidth Mean Shift and Data-Driven Scale Selection , 2001, ICCV.

[5]  James Stephen Marron,et al.  Comparison of data-driven bandwith selectors , 1988 .

[6]  Larry D. Hostetler,et al.  The estimation of the gradient of a density function, with applications in pattern recognition , 1975, IEEE Trans. Inf. Theory.

[7]  Dorin Comaniciu,et al.  Mean shift analysis and applications , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[8]  Keinosuke Fukunaga,et al.  Introduction to statistical pattern recognition (2nd ed.) , 1990 .

[9]  M. Hazelton.  Variable kernel density estimation , 2003 .

[10]  Dorin Comaniciu,et al.  Mean Shift: A Robust Approach Toward Feature Space Analysis , 2002, IEEE Trans. Pattern Anal. Mach. Intell..

[11]  D. W. Scott,et al.  Variable Kernel Density Estimation , 1992 .

[12]  Anil K. Jain,et al.  Data clustering: a review , 1999, CSUR.

[13]  Patrick Pérez,et al.  Detection and segmentation of moving objects in highly dynamic scenes , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[14]  Ian Abramson.  On Bandwidth Variation in Kernel Estimates-A Square Root Law , 1982 .

[15]  C. Quesenberry,et al.  A nonparametric estimate of a multivariate density function , 1965 .

[16]  Yizong Cheng,et al.  Mean Shift, Mode Seeking, and Clustering , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[17]  G. W. Milligan,et al.  An examination of procedures for determining the number of clusters in a data set , 1985 .

[18]  Matthew P. Wand,et al.  Kernel Smoothing , 1995 .

[19]  M. C. Jones,et al.  A reliable data-based bandwidth selection method for kernel density estimation , 1991 .

[20]  J. Marron,et al.  Improved Variable Window Kernel Estimates of Probability Densities , 1995 .

[21]  M. Rosenblatt,et al.  Multivariate k-nearest neighbor density estimates , 1979 .

[22]  J. Simonoff.  Smoothing Methods in Statistics , 1998 .

[23]  J. Simonoff.  Multivariate Density Estimation , 1996 .

[24]  Dorin Comaniciu,et al.  An Algorithm for Data-Driven Bandwidth Selection , 2003, IEEE Trans. Pattern Anal. Mach. Intell..

[25]  Pavel Pudil,et al.  Introduction to Statistical Pattern Recognition , 2006 .

[26]  P. J. Green,et al.  Density Estimation for Statistics and Data Analysis , 1987 .

[27]  Francisco J. Prieto,et al.  Robust covariance matrix estimation and multivariate outlier detection , 1997 .