An Abstraction-Refinement Approach to Verifying Convolutional Neural Networks

Matan Ostrovsky, Clark Barrett, Guy Katz*

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

8 Scopus citations


Convolutional neural networks (CNNs) have achieved immense popularity in areas like computer vision, image processing, speech proccessing, and many others. Unfortunately, despite their excellent performance, they are prone to producing erroneous results — for example, minor perturbations to their inputs can result in severe classification errors. In this paper, we present the Cnn-Abs framework, which implements an abstraction-refinement based scheme for CNN verification. Specifically, Cnn-Abs simplifies the verification problem through the removal of convolutional connections in a way that soundly creates an over-approximation of the original problem; it then iteratively restores these connections if the resulting problem becomes too abstract. Cnn-Abs is designed to use existing verification engines as a backend, and our evaluation demonstrates that it can significantly boost the performance of a state-of-the-art DNN verification engine, reducing runtime by 15.7% on average.

Original languageAmerican English
Title of host publicationAutomated Technology for Verification and Analysis - 20th International Symposium, ATVA 2022, Proceedings
EditorsAhmed Bouajjani, Lukáš Holík, Zhilin Wu
PublisherSpringer Science and Business Media Deutschland GmbH
Number of pages6
ISBN (Print)9783031199912
StatePublished - 2022
Event20th International Symposium on Automated Technology for Verification and Analysis, ATVA 2022 - Virtual, Online
Duration: 25 Oct 202228 Oct 2022

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13505 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference20th International Symposium on Automated Technology for Verification and Analysis, ATVA 2022
CityVirtual, Online

Bibliographical note

Publisher Copyright:
© 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.


Dive into the research topics of 'An Abstraction-Refinement Approach to Verifying Convolutional Neural Networks'. Together they form a unique fingerprint.

Cite this