Abstract
We present a new view of image segmentation by pairwise similarities. We interpret the similarities as edge ows in a Markov random walk and study the eigenvalues and eigenvectors of the walk's transition matrix. This interpretation shows that spectral methods for clustering and segmentation have a probabilistic foundation. In particular, we prove that the Normalized Cut method arises naturally from our framework. Finally, the framework provides a principled method for learning the similarity function as a combination of features. 1 Introduction This paper focuses on pairwise (or similarity-based) clustering and image segmentation. In contrast to statistical clustering methods, that assume a probabilistic model that generates the observed data points (or pixels), pairwise clustering de- nes a similarity function between pairs of points and then formulates a criterion (e.g. maximum total intracluster similarity) that the clustering must optimize. The optimality criteria quant...
Keywords
Affiliated Institutions
Related Publications
Self-organization in vision: stochastic clustering for image segmentation, perceptual grouping, and image database organization
We present a stochastic clustering algorithm which uses pairwise similarity of elements and show how it can be used to address various problems in computer vision, including the...
On Spectral Clustering: Analysis and an algorithm
Despite many empirical successes of spectral clustering methods -- algorithms that cluster points using eigenvectors of matrices derived from the distances between the points --...
Random-Walk Computation of Similarities between Nodes of a Graph with Application to Collaborative Recommendation
This work presents a new perspective on characterizing the similarity between elements of a database or, more generally, nodes of a weighted and undirected graph. It is based on...
Kernel k-means
Kernel k-means and spectral clustering have both been used to identify clusters that are non-linearly separable in input space. Despite significant research, these methods have ...
Segmentation using eigenvectors: a unifying view
Automatic grouping and segmentation of images remains a challenging problem in computer vision. Recently, a number of authors have demonstrated good performance on this task usi...
Publication Info
- Year
- 2000
- Type
- article
- Volume
- 13
- Pages
- 873-879
- Citations
- 370
- Access
- Closed