Nonlinear time series analysis of monkey vocalizations

Abstract

Primate vocalizations are produced as a result of interactions between and within the simple vocal system and the complex signal coming from the nervous system. As a consequence of the nature of this organization, the resulting voice signal is of nonlinear nature. Moreover, in contrast to humans, in many examples of nonhuman primate vocalizations, the vocal folds do not synchorize. Consequently, produced signal is rather complex. Nonlinear techniques were shown to be useful in analyzing nonhuman primate vocalizations. Deterministic versus stochastic (DVS) prediction technique is one of these methods which can be used to determine the amount of nonlinearity measure (LNM), which indicates the presence of a low-dimensional attractor. By using this method, it was demonstrated that while the nonlinearity measure is useful in voice signals with harmonic componente, in highly irregular signals like screams and barks, the detectable amount of nonlinearity was comparatively small. In this study, The amount of nonlinearity in rhesus mankey voices was calculated by using DVS analysis and this measure was used to distinguish different call types and individual properties of the monkeys. Voice signals with harmonic components showed relatively high SNR and low-dimensional nonlinearity, while these phenomena could not be detected in irregular vices. The signals were analyzed and compared among different callers, different call types and also among call subtypes.|Keywords: Deterministic versus Stochastic Analysis, Monkey Vocalization, Rhesus Macaque, Nonlinearity Measure.