How Odontocetes mine an underwater soundscape

Abstract

Toothed whales exceed man-made efforts when detecting and discriminating targets in acoustically challenging environments using bios,onar. They have exquisite abilities to orient themselves by passively listening with acute directional and broad bandwidth hearing a,bilities. We propose to conduct hide-and-seek experiments with two species of trained toothed whales and observe wild specimens sear,ching for fish, to determine important aspects of their behavior during echolocation and soundscape analysis. The combination of exp,eriments and field observations is inspired by studies of bats. We thereby aim at understanding the signals and strategies toothed w,hales use to achieve extraordinary detection and discrimination of targets with complex radiation patterns in cluttered acoustic env,ironments. The odontocete data will be used for deep learning approaches to build relevant models analyzing soundscapes as well as t,o reveal how animals learn and use attention to derive useful cues from their acoustic environment. This data will also allow compar,ison of passive listening schemes in specialist animals to explore differences in strategies when both species explore and behave in, complex soundscapes. We also tap into the neurological learning and decision-making process using data acquisition tags.

Document Details

Document Type
DoD Grant Award
Publication Date
Oct 06, 2022
Source ID
N000142212793

Entities

People

  • Magnus Wahlberg

Organizations

  • Office of Naval Research
  • United States Navy
  • University of Southern Denmark

Tags

Readers

  • Acoustical Oceanography.
  • Marine Mammal Biology
  • Systems Analysis and Design

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks