Discrete Environmental Characterization Using Acoustic Sources of Opportunity and Machine Learning

Abstract

The main objectives of the project are: 1) Develop method to characterize the environment and the associated uncertainties using a" single fixed hydrophone, sources of opportunity, and Bayesian inversion scheme; 2) Formulate and solve the previous problem within a probabilistic ML framework and assess the ML usefulness; 3) Extend the previous developments to an AUV context (moving receiver and limited environmental awareness); 4) Train a MIT/WHOI Joint Program student in both acoustics and ML; 5) Conduct an AUV seatrial to validate the proposed methods. To reach the objectives, the project is divided into technical tasks: data analysis (task 1), development of ML methods (task 2), development of signal processing method (task 3), teaching (task 4) and realization of a field expe"riment (task 5).

Document Details

Document Type
DoD Grant Award
Publication Date
Aug 20, 2019
Source ID
N000141912627

Entities

People

  • Julien Bonnel

Organizations

  • Office of Naval Research
  • United States Navy
  • Woods Hole Oceanographic Institution

Tags

Readers

  • Computational Modeling and Simulation
  • Neural Network Machine Learning.
  • Research Science/Academic Research

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks