A general nonparametric technique is proposed for the analysis of a complex multimodal feature space and to delineate arbitrarily shaped clusters in it. The basic computational module of the technique is an old pattern recognition procedure, the mean shift. We prove for discrete data the convergence of a recursive mean shift procedure to the nearest stationary point of the underlying density function and thus its utility in detecting the modes of the density. The equivalence of the mean shift procedure to the Nadaraya–Watson estimator from kernel regression and the robust M-estimators of location is also established. Algorithms for two low-level vision tasks, discontinuity preserving smoothing and image segmentation are described as applications. In these algorithms the only user set parameter is the resolution of the analysis, and either gray level or color images are accepted as input. Extensive experimental results illustrate their excellent performance.

We present a novel clustering method using the approach of support vector machines. Data points are mapped by means of a Gaussian kernel to a high dimensional feature space, where we search for the minimal enclosing sphere. This sphere, when mapped back to data space, can separate into several components, each enclosing a separate cluster of points. We present a simple algorithm for identifying these clusters. The width of the Gaussian kernel controls the scale at which the data is probed while the soft margin constant helps coping with outliers and overlapping clusters. The structure of a dataset is explored by varying the two parameters, maintaining a minimal number of support vectors to assure smooth cluster boundaries. We demonstrate the performance of our algorithm on several datasets.

A Bayesian-based methodology is presented which automatically penalises over-complex models being fitted to unknown data. We show that, with a Gaussian mixture model, the approach is able to select an `optimal' number of components in the model and so partition data sets. The performance of the Bayesian method is compared to other methods of optimal model selection and found to give good results. The methods are tested on synthetic and real data sets. Introduction Scientific disciplines generate data. In the attempt to understand the patterns present in such data sets methods which perform some form of unsupervised partitioning or modelling are particularly useful. Such an approach is only of use, however, if it offers a less complex representation of the data than the data set itself. This introduces an apparent conflict, however, as any model improves its fit to the data monotonically with increases in its complexity (the number of model parameters) -- a model as complex as the data...

I consider the problem of finding all the modes of a mixture of multivariate Gaussian distributions, which has applications in clustering and regression. I derive exact formulas for the gradient and Hessian and give a partial proof that the number of modes cannot be more than the number of components, and are contained in the convex hull of the component centroids. Then, I develop two exhaustive mode search algorithms: one based on combined quadratic maximisation and gradient ascent and the other one based on a fixed-point iterative scheme. Appropriate values for the search control parameters are derived by taking into account theoretical results regarding the bounds for the gradient and Hessian of the mixture. The significance of the modes is quantified locally (for each mode) by error bars, or confidence intervals (estimated using the values of the Hessian at each mode); and globally by the sparseness of the mixture, measured by its differential entropy (estimated through bounds). I conclude with some reflections about bump-finding.

We present a formulation of boundary mesh segmentation as an optimization problem. Previous segmentation solutions are classified according to the different segmentation goals, the optimization criteria and the various algorithmic techniques used. We identify two primarily distinct types of mesh segmentation, namely parts segmentation and patch segmentation. We also define generic algorithms for the major techniques used for segmentation.

In this work, we present a unified bottom-up and top-down automatic model selection based approach for modelling complex activities of multiple objects in cluttered scenes. An activity of multiple objects is represented based on discrete scene events and their behaviours are modelled by reasoning about the temporal and causal correlations among different events. This is significantly different from the ma-jority of the existing techniques that are centred on object tracking followed by trajectory matching. In our approach, object-independent events are detected and classified by unsupervised clustering us-ing Expectation-Maximisation (EM) and classified using automatic model selection based on Schwarz’s Bayesian Information Criterion (BIC). Dynamic Probabilistic Networks (DPNs) are formulated for mod-elling the temporal and causal correlations among discrete events for robust and holistic scene-level be-haviour interpretation. In particular, we developed a Dynamically Multi-Linked Hidden Markov Model (DML-HMM) based on the discovery of salient dynamic interlinks among multiple temporal processes corresponding to multiple event classes. A DML-HMM is built using BIC based factorisation result-ing in its topology being intrinsically determined by the underlying causality and temporal order among events. Extensive experiments are conducted on modelling activities captured in different indoor and

The extraction of road networks from digital imagery is a fundamental image analysis operation. Common problems encountered in automated road extraction include high sensitivity to typical scene clutter in high-resolution imagery, and Z. Z. inefficiency to meaningfully exploit multispectral imagery MSI . With a ground sample distance GSD of less than 2 m per pixel, roads can be broadly described as elongated regions. We propose an approach of elongated region-based analysis for 2D road extraction from high-resolution imagery, which is suitable for MSI, and is insensitive to conventional edge Z. definition. A self-organising road map SORM algorithm is presented, inspired from a specialised variation of Kohonen's Z. self-organising map SOM neural network algorithm. A spectrally classified high-resolution image is assumed to be the input for our analysis. Our approach proceeds by performing spatial cluster analysis as a mid-level processing technique. This allows us to improve tolerance to road clutter in high-resolution images, and to minimise the effect on road extraction of common classification errors. This approach is designed in consideration of the emerging trend towards high-resolution multispectral sensors. Preliminary results demonstrate robust road extraction ability due to the non-local approach, when presented with noisy input. q 2001 Elsevier Science B.V. All rights reserved.

The mean-shift algorithm, based on ideas proposed by Fukunaga and Hostetler (1975), is a hill-climbing algorithm on the density defined by a finite mixture or a kernel density estimate. Mean-shift can be used as a nonparametric clustering method and has attracted recent attention in computer vision applications such as image segmentation or tracking. We show that, when the kernel is Gaussian, mean-shift is an expectationmaximisation (EM) algorithm, and when the kernel is non-gaussian, mean-shift is a generalised EM algorithm. This implies that mean-shift converges from almost any starting point and that, in general, its convergence is of linear order. For Gaussian mean-shift we show: (1) the rate of linear convergence approaches 0 (superlinear convergence) for very narrow or very wide kernels, but is often close to 1 (thus extremely slow) for intermediate widths, and exactly 1 (sublinear convergence) for widths at which modes merge; (2) the iterates approach the mode along the local principal component of the data points from the inside of the convex hull of the data points; (3) the convergence domains are nonconvex and can be disconnected and show fractal behaviour. We suggest ways of accelerating mean-shift based on the EM interpretation.

We consider a problem intimately related to the creation of maxima under Gaussian blurring: the number of modes of a Gaussian mixture in D dimensions. To our knowledge, a general answer to this question is not known. We conjecture that if the components of the mixture have the same covariance matrix (or the same covariance matrix up to a scaling factor), then the number of modes cannot exceed the number of components. We demonstrate

Problems in data analysis often require the unsupervised partitioning of a data set into clusters. Many methods exist for such partitioning but most have the weakness of being model-based (most assuming hyper-ellipsoidal clusters) or computationally infeasible in anything more than a 3-dimensional data space. We re-consider the notion of cluster analysis in information-theoretic terms and show that minimisation of partition entropy can be used to estimate the number and structure of probable data generators. Keywords: Cluster analysis, data partitioning, information theory. 1 Introduction Many problems in data analysis, especially in signal and image processing, require the unsupervised partitioning of data into a set of `self-similar' clusters or regions. An ideal partition unambiguously assigns each datum to a single cluster and one thinks of the data as being generated by a number of data generators, one for each cluster. Many algorithms have been proposed for such analysis and f...