Akhi A.V. Efficiency of Identification of Complex Noise-Like Signal Classes by Dolphins (<i>Tursiops truncatus</i>) under Simultaneous Presentation Spatial Uncertainty

Abstract

The possibility of dolphin’s auditory system to solve the complex problem of identifying and classifying noise-like signals according to certain invariant features is considered under conditions of spatial uncertainty of their simultaneous presentation. There was studied the ability of bottlenose dolphins, which were trained to recognize and classify such signals, to select a certain class of signals from several that sound simultaneously. The dolphin had to recognize a positive class signal with a pair of simultaneously sounding signals: positive-negative (alternative choice) and with simultaneously sounding three signals: positive-negative-negative (multiple choice). It’s shown that the dolphin effectively solves the problem with a simple alternative choice of two signal sources, at the limit of reliability when choosing from three sources and unreliable when choosing from more sources.