Adversarial training of neural networks using information about activation path differentials

US11657162B2 · kind B2 · utility

Assignee

• Intel Corporation · US

Inventors

• Michael E. Kounavis · Portland, US
• Antonios Papadimitriou · Maplewood, US
• Anindya Paul · Hillsboro, US
• Micah J. Sheller · Hillsboro, US
• Li Chen · Village of La Jolla, US
• Cory Cornelius · Portland, US
• Brandon Edwards · Portland, US

Key dates

• Filing date: Mar 22, 2019
• Grant date: May 23, 2023
• Expiry date: Mar 23, 2042

Classification

• Technology area (CPC G): Physics
• CPC primary: G06N3/044
• WIPO field: Computer technology
• WIPO sector: Electrical engineering

Abstract

In one example an apparatus comprises a memory and a processor to create, from a first deep neural network (DNN) model, a first plurality of DNN models, generate a first set of adversarial examples that are misclassified by the first plurality of deep neural network (DNN) models, determine a first set of activation path differentials between the first plurality of adversarial examples, generate, from the first set of activation path differentials, at least one composite adversarial example which incorporates at least one intersecting critical path that is shared between at least two adversarial examples in the first set of adversarial examples, and use the at least one composite adversarial example to generate a set of inputs for a subsequent training iteration of the DNN model. Other examples may be described.