Final Report: Sparsity-Based Design for Robust Deep Learning-Topic C. iii (3)

Abstract

Deep neural networks [DNNs] yield state of the art performance in many fields, but are known to be vulnerable to small adversarial perturbations. This, together with their lack of interpretability, is a major impediment to their use in many DoD applications, as well as in safety-critical commercial applications such as vehicular autonomy. The overarching goal of this project is to investigate techniques for understanding and robustifying DNNs. The original proposal focused on the specific approach of imposing sparsity constraints to attenuate and eliminate adversarial perturbations, with the goal of obtaining interpretable designs with guaranteed resilience. This is in contrast to the state of art defenses against adversarial perturbations, which are based on black box training with adversarially perturbed examples.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Apr 18, 2022
Accession Number
AD1198333

Entities

People

  • Upamanyu Madhow

Organizations

  • University of California, Santa Barbara

Tags

Communities of Interest

  • Autonomy
  • Energy and Power Technologies
  • Engineered Resilient Systems

DTIC Thesaurus Topics

  • Accuracy
  • Algorithms
  • Artificial Intelligence Software
  • California
  • Carrier Frequencies
  • Computers
  • Convolutional Neural Networks
  • Deep Learning
  • Defense Mechanisms
  • Dimensionality Reduction
  • Engineering
  • Frequency
  • Information Science
  • Information Theory
  • Machine Learning
  • Neural Networks
  • Radio Frequency
  • Signal Processing
  • Standards
  • Students
  • Supervised Machine Learning
  • Transmitters

Fields of Study

  • Computer science

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Distributed Systems and Data Platform Development
  • Neural Network Machine Learning.

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms
  • AI & ML - Neural Networks