Decentralized Perception from Online Learning and Semantic Understanding

Abstract

A multi-state approach to random fields will be investigated for data analysis in a heterogeneous sensing scenario. In order to provide perception of the environment and for situational understanding, we will develop algorithms for sampling of data and scale selection, detection and labeling of states, context building, decentralized modeling and fusion, determining important aspects of data, and prediction. We will demonstrate these algorithms and concepts on embedded platforms in the Duke distributed sensor laboratory.

Document Details

Document Type
DoD Grant Award
Publication Date
Jul 10, 2018
Source ID
N000141812244

Entities

People

  • Vahid Tarokh

Organizations

  • Duke University
  • Office of Naval Research
  • United States Navy

Tags

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Neural Network Machine Learning.
  • Sensor Fusion and Tracking Systems.