Assured Autonomous Vehicle Policy

Assuring Autonomous Airspace Operations

Disentanglement: controlling image attributes during image generation We will develop deep learning generative methods that generate/alter images with respect to specific attributes, while keeping other attributes unchanged, that will help address bias and privacy.

Fairness and Privacy in AI Applied to Healthcare and Autonomy

IDing factors project graphic

Identifying Factors to Explain the Behavior of Deep Learning Systems

Physical Domain Adversarial Machine Learning for Visual Object Recognition

RADICS: Runtime Assurance of Distributed Intelligent Control Systems

Regression Analysis for Autonomy Performance Comparison

Risk-Sensitive Adversarial Learning for Autonomous Systems

Socially Aware Robot Navigation in Human Environments

VALUES: Verified Assured Learning for Unmanned Embedded Systems