Detecting chaos in adversarial examples

The puzzling phenomenon of adversarial examples continues to attract significant research within the machine learning community. The confirmation that adversarial examples can arise in natural real-life circumstances has but increased the interest. While several methods have been proposed for both g...

Full description

Bibliographic Details
Authors: Déniz Suárez, Óscar, Pedraza Dorado, Aníbal, Bueno García, María Gloria
Format: article
Publication Date:2022
Country:España
Institution:Universidad de Castilla-La Mancha
Repository:RUIdeRA. Repositorio Institucional de la UCLM
OAI Identifier:oai:ruidera.uclm.es:10578/45779
Online Access:https://doi.org/10.1016/j.chaos.2022.112577
https://www.sciencedirect.com/science/article/pii/S0960077922007676
https://hdl.handle.net/10578/45779
Access Level:Open access
Keyword:Adversarial examples
Chaos theory
Lyapunov stability
Neural networksDeep learning
Trustworthy machine learning
id ES_19bc26c5bd309cea935d9b95f3deec3e
oai_identifier_str oai:ruidera.uclm.es:10578/45779
network_acronym_str ES
network_name_str España
repository_id_str
spelling Detecting chaos in adversarial examplesDéniz Suárez, ÓscarPedraza Dorado, AníbalBueno García, María GloriaAdversarial examplesChaos theoryLyapunov stabilityNeural networksDeep learningTrustworthy machine learningThe puzzling phenomenon of adversarial examples continues to attract significant research within the machine learning community. The confirmation that adversarial examples can arise in natural real-life circumstances has but increased the interest. While several methods have been proposed for both generating adversarial examples and defending against them, in this work we focus on a previous serendipitous discovery indicating that they can be considered as chaotic signals. More specifically, it has been recently shown that measures of chaoticity in the input signal can be used to detect adversarial examples efficiently. In this work, we extend that approach in two aspects, leading to significant improvements in detection accuracy as demonstrated by results obtained in experiments with four datasets and using seven different attack methods.ELSEVIER202520252022info:eu-repo/semantics/articleapplication/pdfapplication/pdfhttps://doi.org/10.1016/j.chaos.2022.112577https://www.sciencedirect.com/science/article/pii/S0960077922007676https://hdl.handle.net/10578/45779reponame:RUIdeRA. Repositorio Institucional de la UCLMinstname:Universidad de Castilla-La ManchaInglésPDC2021-121197-C22info:eu-repo/semantics/openAccessoai:ruidera.uclm.es:10578/457792026-05-27T07:36:41Z
dc.title.none.fl_str_mv Detecting chaos in adversarial examples
title Detecting chaos in adversarial examples
spellingShingle Detecting chaos in adversarial examples
Déniz Suárez, Óscar
Adversarial examples
Chaos theory
Lyapunov stability
Neural networksDeep learning
Trustworthy machine learning
title_short Detecting chaos in adversarial examples
title_full Detecting chaos in adversarial examples
title_fullStr Detecting chaos in adversarial examples
title_full_unstemmed Detecting chaos in adversarial examples
title_sort Detecting chaos in adversarial examples
dc.creator.none.fl_str_mv Déniz Suárez, Óscar
Pedraza Dorado, Aníbal
Bueno García, María Gloria
author Déniz Suárez, Óscar
author_facet Déniz Suárez, Óscar
Pedraza Dorado, Aníbal
Bueno García, María Gloria
author_role author
author2 Pedraza Dorado, Aníbal
Bueno García, María Gloria
author2_role author
author
dc.subject.none.fl_str_mv Adversarial examples
Chaos theory
Lyapunov stability
Neural networksDeep learning
Trustworthy machine learning
topic Adversarial examples
Chaos theory
Lyapunov stability
Neural networksDeep learning
Trustworthy machine learning
description The puzzling phenomenon of adversarial examples continues to attract significant research within the machine learning community. The confirmation that adversarial examples can arise in natural real-life circumstances has but increased the interest. While several methods have been proposed for both generating adversarial examples and defending against them, in this work we focus on a previous serendipitous discovery indicating that they can be considered as chaotic signals. More specifically, it has been recently shown that measures of chaoticity in the input signal can be used to detect adversarial examples efficiently. In this work, we extend that approach in two aspects, leading to significant improvements in detection accuracy as demonstrated by results obtained in experiments with four datasets and using seven different attack methods.
publishDate 2022
dc.date.none.fl_str_mv 2022
2025
2025
dc.type.none.fl_str_mv info:eu-repo/semantics/article
format article
dc.identifier.none.fl_str_mv https://doi.org/10.1016/j.chaos.2022.112577
https://www.sciencedirect.com/science/article/pii/S0960077922007676
https://hdl.handle.net/10578/45779
url https://doi.org/10.1016/j.chaos.2022.112577
https://www.sciencedirect.com/science/article/pii/S0960077922007676
https://hdl.handle.net/10578/45779
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.relation.none.fl_str_mv PDC2021-121197-C22
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
application/pdf
dc.publisher.none.fl_str_mv ELSEVIER
publisher.none.fl_str_mv ELSEVIER
dc.source.none.fl_str_mv reponame:RUIdeRA. Repositorio Institucional de la UCLM
instname:Universidad de Castilla-La Mancha
instname_str Universidad de Castilla-La Mancha
reponame_str RUIdeRA. Repositorio Institucional de la UCLM
collection RUIdeRA. Repositorio Institucional de la UCLM
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869404063519473664
score 15,811543

Similar Items