PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Class distributions on SOM surfaces for feature extraction and object retrieval
Jorma Laaksonen, Markus Koskela and Erkki Oja
Neural Networks Volume 17, Number 8-9, pp. 1121-1133, 2004.

This is the latest version of this eprint.


A Self-Organizing Map (SOM) is typically trained in unsupervised mode, using a large batch of training data. If the data contain semantically related object groupings or classes, subsets of vectors belonging to such user-defined classes can be mapped on the SOM by finding the best matching unit for each vector in the set. The distribution of the data vectors over the map forms a two-dimensional discrete probability density. Even from the same data, qualitatively different distributions can be obtained by using different feature extraction techniques. We used such feature distributions for comparing different classes and different feature representations of the data in the context of our content-based image retrieval system PicSOM. The information-theoretic measures of entropy and mutual information are suggested to evaluate the compactness of a distribution and the independence of two distributions. Also, the effect of low-pass filtering the SOM surfaces prior to the calculation of the entropy is studied.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Article
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Computational, Information-Theoretic Learning with Statistics
Machine Vision
Information Retrieval & Textual Information Access
ID Code:2546
Deposited By:Erkki Oja
Deposited On:22 November 2006

Available Versions of this Item