AUDIO SIGNAL BASED ENVIORNMENT DETERMINATION FOR MOBILE ROBOTS BY TIME-FREQUENCY ANALYSIS

Abstract

Audio feature extraction and classification are important process for audio signal analysis in many applications, such as content-based retrieval, multimedia database, and auditory scene analysis. Efficient feature extraction is required by these applications. To determine the information of environments visually is an important problem for mobile robots. In this paper, we describe statistical features of different sound signals were extracted by using time frequency analysis to classify them by using neural network. In this work, classification of four signals was achieved successfully with classification score up to 0.97.