Magnasco, M. | Search bioRxiv

bioRxiv2019

Performance Evaluation of Traditional Signal Processing Methods in Localizing Tursiops truncatus Whistles in a Reverberant Aquatic Environment

Relative to individually distinctive signature whistles, little is known about the \"non-signature\" calls - particularly the non-signature whistles - of the common Atlantic bottlenose dolphin, Tursiops truncatus. While such calls are suspected to serve social function, tracking their exchange among conspecifics and correlating their usage with non-acoustic behavior has proven challenging, given both their relative scarcity in the dolphin repertoire and their characteristic shared use among dolphins, which precludes the unique identification of callers on the basis of whistle properties alone. Towards the goal of robustly identifying the callers of non-signature whistles (equivalently, attributing non-signature whistles to callers), we present a new, long-term audiovisual monitoring system designed for and tested at the Dolphin Discovery exhibit of the National Aquarium in Baltimore, Maryland. In this paper, we confirm the systems ability to spatially localize impulse-like sounds using traditional signal processing approaches that have already been used to localize dolphin echolocation clicks. We go on to provide the first rigorous experimental evaluation of the component time-difference-of-arrival-(TDOA) extraction methods on whistle-like tonal sounds in a (reverberant) aquatic environment, showing that they are generally not suited to sound localization. Nevertheless, we find that TDOA extraction under these circumstances is performed significantly better using a Generalized Cross-Correlation with Phase Transform (GCC-PHAT) method than a standard circular cross-correlation method, a potentially important result.

biophysics↗

bioRxiv2019

Machine Source Localization of Tursiops Truncatus Whistle-like Sounds in a Reverberant Aquatic Environment

Most research into bottlenose dolphins (Tursiops truncatus) capacity for communication has centered on tonal calls termed whistles, in particular individually distinctive contact calls referred to as signature whistles. While "non-signature" whistles exist, and may be important components of bottlenose dolphins communicative repertoire, they have not been studied extensively. This is in part due to the difficulty of attributing whistles to specific individuals, a challenge that has limited the study of not only non-signature whistles but the study of general acoustic exchanges among socializing dolphins. In this paper, we propose the first machine-learning-based approach to identifying the source locations of semi-stationary, tonal, whistle-like sounds in a highly reverberant space, specifically a half-cylindrical dolphin pool. We deliver estimated time-difference-of-arrivals (TDOAs) and normalized cross-correlation values computed from pairs of hydrophone signals to a random forest model for high-feature-volume classification and feature selection, and subsequently deliver the selected features into linear discriminant analysis, linear and quadratic Support Vector Machine (SVM), and Gaussian process models. In our 14-source-location setup, we achieve perfect accuracy in localization by classification and high accuracy in localization by regression (median absolute deviation of 0.66 m, interquartile range of 0.34 m - 1.57 m), with fewer than 10,000 features. By building a parsimonious (minimum-feature) classification tree model for the same task, we show that a minimally sufficient feature set is consistent with the information valued by a strictly geometric, time-difference-of-arrival-based approach to sound source localization. Ultimately, our regression models yielded better accuracy than the established Steered-Response Power (SRP) method when all training data were used, and comparable accuracy along the pool surface when deprived of training data at testing sites; our methods additionally boast improved computation time and the potential for superior localization accuracy in all dimensions with more training data.

biophysics↗