Verification of Deep Convolutional Neural Networks Using ImageStars

Hoang Dung Tran, Stanley Bak, Weiming Xiang, Taylor T. Johnson

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Scopus citations

Abstract

Convolutional Neural Networks (CNN) have redefined state-of-the-art in many real-world applications, such as facial recognition, image classification, human pose estimation, and semantic segmentation. Despite their success, CNNs are vulnerable to adversarial attacks, where slight changes to their inputs may lead to sharp changes in their output in even well-trained networks. Set-based analysis methods can detect or prove the absence of bounded adversarial attacks, which can then be used to evaluate the effectiveness of neural network training methodology. Unfortunately, existing verification approaches have limited scalability in terms of the size of networks that can be analyzed. In this paper, we describe a set-based framework that successfully deals with real-world CNNs, such as VGG16 and VGG19, that have high accuracy on ImageNet. Our approach is based on a new set representation called the ImageStar, which enables efficient exact and over-approximative analysis of CNNs. ImageStars perform efficient set-based analysis by combining operations on concrete images with linear programming (LP). Our approach is implemented in a tool called NNV, and can verify the robustness of VGG networks with respect to a small set of input states, derived from adversarial attacks, such as the DeepFool attack. The experimental results show that our approach is less conservative and faster than existing zonotope and polytope methods.

Original languageEnglish (US)
Title of host publicationComputer Aided Verification - 32nd International Conference, CAV 2020, Proceedings
EditorsShuvendu K. Lahiri, Chao Wang
PublisherSpringer
Pages18-42
Number of pages25
ISBN (Print)9783030532871
DOIs
StatePublished - 2020
Event32nd International Conference on Computer Aided Verification, CAV 2020 - Los Angeles, United States
Duration: Jul 21 2020Jul 24 2020

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12224 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference32nd International Conference on Computer Aided Verification, CAV 2020
CountryUnited States
CityLos Angeles
Period7/21/207/24/20

Keywords

  • Computer vision
  • Machine learning
  • Neural networks
  • Reachability analysis

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Verification of Deep Convolutional Neural Networks Using ImageStars'. Together they form a unique fingerprint.

  • Cite this

    Tran, H. D., Bak, S., Xiang, W., & Johnson, T. T. (2020). Verification of Deep Convolutional Neural Networks Using ImageStars. In S. K. Lahiri, & C. Wang (Eds.), Computer Aided Verification - 32nd International Conference, CAV 2020, Proceedings (pp. 18-42). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12224 LNCS). Springer. https://doi.org/10.1007/978-3-030-53288-8_2