Habituation based on spectrogram analysis

Abstract

In this paper we present a habituation mechanism which includes a modification of the Stanley's habituation model with the addition of a stage based on spectrogram to detect temporal patterns in a signal and to obtain a measure of habituation to these patterns. This means that this measure shows a saturation process as the pattern is perceived by the system and when it disappears the measure drops. The use of the spectrogram simplifies the detection of the temporal patterns which can be detected with naive techniques. We have carried on some experiments both a synthetic signal and real signals like readings of a sonar in a mobile robot.