Forthepurposeofdevelopingareal-timetransportationnoiserecognitionsystem,avarietyof acousticalfeaturesandstatisticalmodelsof signalsare reviewed.Findingan appropriateacousticalcriteria to discriminatebetweenseveraltypes of environmentalnoise is the goal of the present paper. The main emphasisis put on the discriminationof transportationnoise,musicand speech. The resultscouldbe used in transporta...

A framework for the classi cation of features with respect to scale is proposed where the visual impression of features is linearly combined at dif ferent scales The framework is applied to typical feature extraction tasks as a tool for the identi cation of images of similar visual content A possi ble formulation of queries for retrieving images by primitive visual content taking scale into acc...

This paper shows the importance of the use of class information in feature extraction for classification and inappropriateness of conventional PCA to feature extraction for classification. We consider two eigenvector-based approaches that take into account the class information. The first approach is parametric and optimizes the ratio of between-class variance to within-class variance of the tr...

Article history: Received 16 May 2008 Received in revised form 24 October 2008 Accepted 7 January 2009

In this paper, we present a robot architecture that enables automatic mapping in unmodified office buildings during normal business hours. The architecture that we describe, which is an instance of the feature-extraction approach to mapping and navigation, is also unique in its ability to hypothesize the existence of unseen passages in its environment and then test and validate its hypotheses t...