Bayesian approach for object detection and scene analysis
Abstract
Despite impressive progress in machine learning techniques over the last decade, current state of the art (SOA) models fall short of the capabilities of human vision. In contrast to human vision, current SOA models require massive amounts of training data, are not robust to even imperceptible changes in inputs, and are hard to interpret. This project aims to develop computer vision algorithms and network structures that will be able to capture important properties of human vision such as the ability to integrate contextual information during the recognition process and learn from small numbers of training examples.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 24, 2025
- Source ID
- N000142512132
Entities
People
- Alan Yuille
Organizations
- Johns Hopkins University
- Office of Naval Research
- United States Navy