Visual scene understanding: A Bayesian approach

Abstract

Many methods for visual scene understanding rely on supervised training using very large data sets. While these methods can be succe"ssful in certain circumstances, they are very costly, brittle and provide little genuine understanding. Instead, we will develop a c"omputational framework for autonomous scene analysis that will learn object categories from a small number of observations. The proposed effort will use a Bayesian approach to learn object categories and relations between objects and object parts. Our approach will develop a framework for using both spatial and temporal contextual information and for integrating the two types of information during the learning and inference stages.

Document Details

Document Type
DoD Grant Award
Publication Date
May 05, 2017
Source ID
N000141712509

Entities

People

  • Ernst Niebur

Organizations

  • Johns Hopkins University
  • Office of Naval Research
  • United States Navy

Tags

Fields of Study

  • Computer science

Readers

  • Educational Psychology
  • Neural Network Machine Learning.

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks