ManiGen: A Manifold Aided Black-Box Generator of Adversarial Examples
From recent research work, it has been shown that neural network (NN) classifiers are vulnerable to adversarial examples which contain special perturbations that are ignored by human eyes while can mislead NN classifiers. In this paper, we propose a practical black-box adversarial example generator,...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2020-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9216163/ |
| _version_ | 1818662448765337600 |
|---|---|
| author | Guanxiong Liu Issa Khalil Abdallah Khreishah Abdulelah Algosaibi Adel Aldalbahi Mohammed Alnaeem Abdulaziz Alhumam Muhammad Anan |
| author_facet | Guanxiong Liu Issa Khalil Abdallah Khreishah Abdulelah Algosaibi Adel Aldalbahi Mohammed Alnaeem Abdulaziz Alhumam Muhammad Anan |
| author_sort | Guanxiong Liu |
| collection | DOAJ |
| description | From recent research work, it has been shown that neural network (NN) classifiers are vulnerable to adversarial examples which contain special perturbations that are ignored by human eyes while can mislead NN classifiers. In this paper, we propose a practical black-box adversarial example generator, dubbed ManiGen. ManiGen does not require any knowledge of the inner state of the target classifier. It generates adversarial examples by searching along the manifold, which is a concise representation of input data. Through extensive set of experiments on different datasets, we show that (1) adversarial examples generated by ManiGen can mislead standalone classifiers by being as successful as the state-of-the-art white-box generator, Carlini, and (2) adversarial examples generated by ManiGen can more effectively attack classifiers with state-of-the-art defenses. |
| first_indexed | 2024-12-17T05:01:07Z |
| format | Article |
| id | doaj.art-94eedc410fbe46468321f9922f968c27 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-17T05:01:07Z |
| publishDate | 2020-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-94eedc410fbe46468321f9922f968c272022-12-21T22:02:34ZengIEEEIEEE Access2169-35362020-01-01819708619709610.1109/ACCESS.2020.30292709216163ManiGen: A Manifold Aided Black-Box Generator of Adversarial ExamplesGuanxiong Liu0https://orcid.org/0000-0001-7620-5836Issa Khalil1https://orcid.org/0000-0002-7660-9512Abdallah Khreishah2https://orcid.org/0000-0003-1583-713XAbdulelah Algosaibi3Adel Aldalbahi4https://orcid.org/0000-0003-1903-0480Mohammed Alnaeem5https://orcid.org/0000-0002-5682-6237Abdulaziz Alhumam6https://orcid.org/0000-0001-7778-2838Muhammad Anan7New Jersey Institute of Technology, Newark, NJ, USAQatar Computing Research Institute, Doha, QatarNew Jersey Institute of Technology, Newark, NJ, USAKing Faisal University, Al-Hasa, Saudi ArabiaKing Faisal University, Al-Hasa, Saudi ArabiaKing Faisal University, Al-Hasa, Saudi ArabiaKing Faisal University, Al-Hasa, Saudi ArabiaAlfaisal University, Riyadh, Saudi ArabiaFrom recent research work, it has been shown that neural network (NN) classifiers are vulnerable to adversarial examples which contain special perturbations that are ignored by human eyes while can mislead NN classifiers. In this paper, we propose a practical black-box adversarial example generator, dubbed ManiGen. ManiGen does not require any knowledge of the inner state of the target classifier. It generates adversarial examples by searching along the manifold, which is a concise representation of input data. Through extensive set of experiments on different datasets, we show that (1) adversarial examples generated by ManiGen can mislead standalone classifiers by being as successful as the state-of-the-art white-box generator, Carlini, and (2) adversarial examples generated by ManiGen can more effectively attack classifiers with state-of-the-art defenses.https://ieeexplore.ieee.org/document/9216163/Adversarial examplesmachine learningneural networkmanifold |
| spellingShingle | Guanxiong Liu Issa Khalil Abdallah Khreishah Abdulelah Algosaibi Adel Aldalbahi Mohammed Alnaeem Abdulaziz Alhumam Muhammad Anan ManiGen: A Manifold Aided Black-Box Generator of Adversarial Examples IEEE Access Adversarial examples machine learning neural network manifold |
| title | ManiGen: A Manifold Aided Black-Box Generator of Adversarial Examples |
| title_full | ManiGen: A Manifold Aided Black-Box Generator of Adversarial Examples |
| title_fullStr | ManiGen: A Manifold Aided Black-Box Generator of Adversarial Examples |
| title_full_unstemmed | ManiGen: A Manifold Aided Black-Box Generator of Adversarial Examples |
| title_short | ManiGen: A Manifold Aided Black-Box Generator of Adversarial Examples |
| title_sort | manigen a manifold aided black box generator of adversarial examples |
| topic | Adversarial examples machine learning neural network manifold |
| url | https://ieeexplore.ieee.org/document/9216163/ |
| work_keys_str_mv | AT guanxiongliu manigenamanifoldaidedblackboxgeneratorofadversarialexamples AT issakhalil manigenamanifoldaidedblackboxgeneratorofadversarialexamples AT abdallahkhreishah manigenamanifoldaidedblackboxgeneratorofadversarialexamples AT abdulelahalgosaibi manigenamanifoldaidedblackboxgeneratorofadversarialexamples AT adelaldalbahi manigenamanifoldaidedblackboxgeneratorofadversarialexamples AT mohammedalnaeem manigenamanifoldaidedblackboxgeneratorofadversarialexamples AT abdulazizalhumam manigenamanifoldaidedblackboxgeneratorofadversarialexamples AT muhammadanan manigenamanifoldaidedblackboxgeneratorofadversarialexamples |