Unsupervised scene classification based on context of features for a mobile robot

Abstract

This paper presents an unsupervised scene classification method based on the context of features for semantic recognition of indoor scenes used for an autonomous mobile robot. Our method creates Visual Words (VWs) of two types using Scale-Invariant Feature Transform (SIFT) and Gist. Using the combination of VWs, our method creates Bags of VWs (BoVWs) to vote for a two-dimensional histogram as context-based features. Moreover, our method generates labels as a candidate of categories while maintaining stability and plasticity together using the incremental learning function of Adaptive Resonance Theory-2 (ART-2). Our method actualizes unsupervised-learning-based scene classification using generated labels of ART-2 as teaching signals of Counter Propagation Networks (CPNs). The spatial and topological relations among scenes are mapped on the category map of CPNs. The relations of classified scenes that include categories are visualized on the category map. The experiment demonstrates the classification accuracy of semantic categories such as office rooms and corridors using an open dataset as an evaluation platform of position estimation and navigation for an autonomous mobile robot. © 2011 Springer-Verlag.