Adversarial Attack Type I: Cheat Classifiers by Significant Changes

Despite the great success of deep neural networks, the adversarial attack can cheat some well-trained classifiers by small permutations. In this paper, we propose another type of adversarial attack that can cheat classifiers by significant changes. For example, we can significantly change a face but...

Celý popis

Uložené v:

Podrobná bibliografia
Vydané v:	IEEE transactions on pattern analysis and machine intelligence Ročník 43; číslo 3; s. 1100 - 1109
Hlavní autori:	Tang, Sanli, Huang, Xiaolin, Chen, Mingjian, Sun, Chengjin, Yang, Jie
Médium:	Journal Article
Jazyk:	English
Vydavateľské údaje:	United States IEEE 01.03.2021 The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Predmet:	Adversarial attack Aerospace electronics Artificial neural networks Classifiers Face recognition Neural networks Permutations Sun supervised variational autoencoder Task analysis Toy manufacturing industry Training type I error
ISSN:	0162-8828, 1939-3539, 2160-9292, 1939-3539
On-line prístup:	Získať plný text
Tagy:	Pridať tag Žiadne tagy, Buďte prvý, kto otaguje tento záznam!

Abstract	Despite the great success of deep neural networks, the adversarial attack can cheat some well-trained classifiers by small permutations. In this paper, we propose another type of adversarial attack that can cheat classifiers by significant changes. For example, we can significantly change a face but well-trained neural networks still recognize the adversarial and the original example as the same person. Statistically, the existing adversarial attack increases Type II error and the proposed one aims at Type I error, which are hence named as Type II and Type I adversarial attack, respectively. The two types of attack are equally important but are essentially different, which are intuitively explained and numerically evaluated. To implement the proposed attack, a supervised variation autoencoder is designed and then the classifier is attacked by updating the latent variables using gradient information. Besides, with pre-trained generative models, Type I attack on latent spaces is investigated as well. Experimental results show that our method is practical and effective to generate Type I adversarial examples on large-scale image datasets. Most of these generated examples can pass detectors designed for defending Type II attack and the strengthening strategy is only efficient with a specific type attack, both implying that the underlying reasons for Type I and Type II attack are different.
AbstractList	Despite the great success of deep neural networks, the adversarial attack can cheat some well-trained classifiers by small permutations. In this paper, we propose another type of adversarial attack that can cheat classifiers by significant changes. For example, we can significantly change a face but well-trained neural networks still recognize the adversarial and the original example as the same person. Statistically, the existing adversarial attack increases Type II error and the proposed one aims at Type I error, which are hence named as Type II and Type I adversarial attack, respectively. The two types of attack are equally important but are essentially different, which are intuitively explained and numerically evaluated. To implement the proposed attack, a supervised variation autoencoder is designed and then the classifier is attacked by updating the latent variables using gradient information. Besides, with pre-trained generative models, Type I attack on latent spaces is investigated as well. Experimental results show that our method is practical and effective to generate Type I adversarial examples on large-scale image datasets. Most of these generated examples can pass detectors designed for defending Type II attack and the strengthening strategy is only efficient with a specific type attack, both implying that the underlying reasons for Type I and Type II attack are different.Despite the great success of deep neural networks, the adversarial attack can cheat some well-trained classifiers by small permutations. In this paper, we propose another type of adversarial attack that can cheat classifiers by significant changes. For example, we can significantly change a face but well-trained neural networks still recognize the adversarial and the original example as the same person. Statistically, the existing adversarial attack increases Type II error and the proposed one aims at Type I error, which are hence named as Type II and Type I adversarial attack, respectively. The two types of attack are equally important but are essentially different, which are intuitively explained and numerically evaluated. To implement the proposed attack, a supervised variation autoencoder is designed and then the classifier is attacked by updating the latent variables using gradient information. Besides, with pre-trained generative models, Type I attack on latent spaces is investigated as well. Experimental results show that our method is practical and effective to generate Type I adversarial examples on large-scale image datasets. Most of these generated examples can pass detectors designed for defending Type II attack and the strengthening strategy is only efficient with a specific type attack, both implying that the underlying reasons for Type I and Type II attack are different. Despite the great success of deep neural networks, the adversarial attack can cheat some well-trained classifiers by small permutations. In this paper, we propose another type of adversarial attack that can cheat classifiers by significant changes. For example, we can significantly change a face but well-trained neural networks still recognize the adversarial and the original example as the same person. Statistically, the existing adversarial attack increases Type II error and the proposed one aims at Type I error, which are hence named as Type II and Type I adversarial attack, respectively. The two types of attack are equally important but are essentially different, which are intuitively explained and numerically evaluated. To implement the proposed attack, a supervised variation autoencoder is designed and then the classifier is attacked by updating the latent variables using gradient information. Besides, with pre-trained generative models, Type I attack on latent spaces is investigated as well. Experimental results show that our method is practical and effective to generate Type I adversarial examples on large-scale image datasets. Most of these generated examples can pass detectors designed for defending Type II attack and the strengthening strategy is only efficient with a specific type attack, both implying that the underlying reasons for Type I and Type II attack are different.
Author	Tang, Sanli Huang, Xiaolin Yang, Jie Chen, Mingjian Sun, Chengjin
Author_xml	– sequence: 1 givenname: Sanli surname: Tang fullname: Tang, Sanli email: tangsanli@sjtu.edu.cn organization: MOE Key Laboratory of System Control and Information Processing, Institute of Image Processing and Pattern Recognition and Institute of Medical Robotics, Shanghai Jiao Tong University, Shanghai, P.R. China – sequence: 2 givenname: Xiaolin orcidid: 0000-0003-4285-6520 surname: Huang fullname: Huang, Xiaolin email: xiaolinhuang@sjtu.edu.cn organization: MOE Key Laboratory of System Control and Information Processing, Institute of Image Processing and Pattern Recognition and Institute of Medical Robotics, Shanghai Jiao Tong University, Shanghai, P.R. China – sequence: 3 givenname: Mingjian orcidid: 0000-0003-0584-6286 surname: Chen fullname: Chen, Mingjian email: w179261466@sjtu.edu.cn organization: MOE Key Laboratory of System Control and Information Processing, Institute of Image Processing and Pattern Recognition and Institute of Medical Robotics, Shanghai Jiao Tong University, Shanghai, P.R. China – sequence: 4 givenname: Chengjin orcidid: 0000-0002-9992-7919 surname: Sun fullname: Sun, Chengjin email: sunchengjin@sjtu.edu.cn organization: MOE Key Laboratory of System Control and Information Processing, Institute of Image Processing and Pattern Recognition and Institute of Medical Robotics, Shanghai Jiao Tong University, Shanghai, P.R. China – sequence: 5 givenname: Jie orcidid: 0000-0003-4801-7162 surname: Yang fullname: Yang, Jie email: jieyang@sjtu.edu.cn organization: MOE Key Laboratory of System Control and Information Processing, Institute of Image Processing and Pattern Recognition and Institute of Medical Robotics, Shanghai Jiao Tong University, Shanghai, P.R. China
BackLink	https://www.ncbi.nlm.nih.gov/pubmed/31442970$$D View this record in MEDLINE/PubMed
BookMark	eNp9kV1LwzAUhoMoOqd_QEEK3njTmZw0TeLdKH4MFAXndTjNUo127Uw6Yf_e6qYXXgiBEPI8L4fz7pPtpm0cIUeMjhij-nz6ML6bjIAyPQLNcy7VFhkAy2mqQcM2GVCWQ6oUqD2yH-MrpSwTlO-SPc6yDLSkA1KMZx8uRAwe62TcdWjfkulq4ZLJRVK8OOySosYYfeV7KilXyaN_bvqXxab_esHm2cUDslNhHd3h5h6Sp6vLaXGT3t5fT4rxbWq5YF0q0Zal1gJsplCUbJZbKzOArBIWS9BWuBw5KuCOQ45aKp1XloGtpKIaSz4kZ-vcRWjfly52Zu6jdXWNjWuX0QAoKgQInffo6R_0tV2Gpp_OQKZkxiXvz5CcbKhlOXczswh-jmFlftbTA2oN2NDGGFxlrO-w823TBfS1YdR8NWG-mzBfTZhNE70Kf9Sf9H-l47XknXO_glJUcib4Jz_kkeg
CODEN	ITPIDJ
CitedBy_id	crossref_primary_10_1016_j_patcog_2021_108491 crossref_primary_10_1016_j_patcog_2021_108192 crossref_primary_10_1007_s11263_025_02467_7 crossref_primary_10_1109_TAI_2023_3257276 crossref_primary_10_1109_TPAMI_2025_3540200 crossref_primary_10_1109_TDSC_2022_3186918 crossref_primary_10_1109_TMI_2021_3072568 crossref_primary_10_1109_TETCI_2024_3367812 crossref_primary_10_1109_TPAMI_2020_3033291 crossref_primary_10_1016_j_engappai_2023_106595 crossref_primary_10_3390_technologies13050202 crossref_primary_10_1145_3643563 crossref_primary_10_1109_TIP_2025_3572793 crossref_primary_10_32604_cmc_2022_029969 crossref_primary_10_1109_TPAMI_2024_3365699 crossref_primary_10_1109_JIOT_2020_3040281
Cites_doi	10.1109/TNNLS.2018.2886017 10.1109/5.726791 10.4204/EPTCS.257.3 10.1109/CVPR.2017.645 10.24963/ijcai.2018/543 10.1109/SP.2016.41 10.1109/SP.2017.49 10.1201/9781351251389-8 10.1109/CVPR.2016.282 10.1109/CVPR.2019.00453 10.1109/ICCV.2015.425 10.14722/ndss.2018.23198 10.1109/CVPR.2017.660 10.1109/CVPR.2016.90 10.1109/CVPR.2015.7298682 10.1109/CVPR.2015.7298640
ContentType	Journal Article
Copyright	Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021
Copyright_xml	– notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021
DBID	97E RIA RIE AAYXX CITATION NPM 7SC 7SP 8FD JQ2 L7M L~C L~D 7X8
DOI	10.1109/TPAMI.2019.2936378
DatabaseName	IEEE Xplore (IEEE) IEEE All-Society Periodicals Package (ASPP) 1998–Present IEEE Electronic Library (IEL) CrossRef PubMed Computer and Information Systems Abstracts Electronics & Communications Abstracts Technology Research Database ProQuest Computer Science Collection Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Academic Computer and Information Systems Abstracts Professional MEDLINE - Academic
DatabaseTitle	CrossRef PubMed Technology Research Database Computer and Information Systems Abstracts – Academic Electronics & Communications Abstracts ProQuest Computer Science Collection Computer and Information Systems Abstracts Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Professional MEDLINE - Academic
DatabaseTitleList	MEDLINE - Academic Technology Research Database PubMed
Database_xml	– sequence: 1 dbid: NPM name: PubMed url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed sourceTypes: Index Database – sequence: 2 dbid: RIE name: IEEE Electronic Library (IEL) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher – sequence: 3 dbid: 7X8 name: MEDLINE - Academic url: https://search.proquest.com/medline sourceTypes: Aggregation Database
DeliveryMethod	fulltext_linktorsrc
Discipline	Engineering Computer Science
EISSN	2160-9292 1939-3539
EndPage	1109
ExternalDocumentID	31442970 10_1109_TPAMI_2019_2936378 8807315
Genre	orig-research Journal Article
GrantInformation_xml	– fundername: Committee of Science and Technology, Shanghai, China grantid: 19510711200 – fundername: National Natural Science Foundation of China grantid: 61977046; 61603248; 61876107; U1803261 funderid: 10.13039/501100001809 – fundername: 1000-Talent Plan
GroupedDBID	--- -DZ -~X .DC 0R~ 29I 4.4 53G 5GY 6IK 97E AAJGR AARMG AASAJ AAWTH ABAZT ABQJQ ABVLG ACGFO ACGFS ACIWK ACNCT AENEX AGQYO AHBIQ AKJIK AKQYR ALMA_UNASSIGNED_HOLDINGS ASUFR ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 DU5 E.L EBS EJD F5P HZ~ IEDLZ IFIPE IPLJI JAVBF LAI M43 MS~ O9- OCL P2P PQQKQ RIA RIE RNS RXW TAE TN5 UHB ~02 AAYXX CITATION 5VS 9M8 AAYOK ABFSI ADRHT AETIX AGSQL AI. AIBXA ALLEH FA8 H~9 IBMZZ ICLAB IFJZH NPM PKN RIC RIG RNI RZB VH1 XJT Z5M 7SC 7SP 8FD JQ2 L7M L~C L~D 7X8
ID	FETCH-LOGICAL-c351t-7acbb9952c48a5b1d6cc74224f5cab29c5e6a3a823e326a97896fc12cf7809ab3
IEDL.DBID	RIE
ISICitedReferencesCount	19
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000616309900025&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	0162-8828 1939-3539
IngestDate	Sun Sep 28 02:49:43 EDT 2025 Sun Nov 09 06:24:13 EST 2025 Wed Feb 19 02:29:47 EST 2025 Sat Nov 29 05:15:59 EST 2025 Tue Nov 18 22:41:42 EST 2025 Wed Aug 27 05:47:02 EDT 2025
IsPeerReviewed	true
IsScholarly	true
Issue	3
Language	English
License	https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html https://doi.org/10.15223/policy-029 https://doi.org/10.15223/policy-037
LinkModel	DirectLink
MergedId	FETCHMERGED-LOGICAL-c351t-7acbb9952c48a5b1d6cc74224f5cab29c5e6a3a823e326a97896fc12cf7809ab3
Notes	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 content type line 23
ORCID	0000-0003-4285-6520 0000-0003-0584-6286 0000-0003-4801-7162 0000-0002-9992-7919
PMID	31442970
PQID	2487437337
PQPubID	85458
PageCount	10
ParticipantIDs	crossref_citationtrail_10_1109_TPAMI_2019_2936378 pubmed_primary_31442970 crossref_primary_10_1109_TPAMI_2019_2936378 proquest_journals_2487437337 proquest_miscellaneous_2280552596 ieee_primary_8807315
PublicationCentury	2000
PublicationDate	2021-03-01
PublicationDateYYYYMMDD	2021-03-01
PublicationDate_xml	– month: 03 year: 2021 text: 2021-03-01 day: 01
PublicationDecade	2020
PublicationPlace	United States
PublicationPlace_xml	– name: United States – name: New York
PublicationTitle	IEEE transactions on pattern analysis and machine intelligence
PublicationTitleAbbrev	TPAMI
PublicationTitleAlternate	IEEE Trans Pattern Anal Mach Intell
PublicationYear	2021
Publisher	IEEE The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml	– name: IEEE – name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References	ref35 szegedy (ref9) 2014 ref12 goodfellow (ref6) 2015 ref36 kingma (ref34) 2015 ref31 ref11 ref32 chen (ref13) 2018 ref1 ref39 krizhevsky (ref2) 2012 baluja (ref19) 2018 huang (ref38) 2007 wang (ref10) 2017 ref18 huang (ref22) 2018 madry (ref14) 2018 lample (ref30) 2017 dai (ref29) 2017 ref23 ref26 ref25 ref41 ref21 song (ref15) 2018 kannan (ref40) 2018 kingma (ref5) 2014 makhzani (ref16) 2015 kingma (ref28) 2014 berthelot (ref33) 2017 ref8 ref7 lee (ref20) 2017 ref3 goodfellow (ref4) 2014 tramer (ref42) 2018 mirza (ref27) 2014 odena (ref17) 2017 gilmer (ref24) 2018 abadi (ref37) 2016
References_xml	– ident: ref8 doi: 10.1109/TNNLS.2018.2886017 – ident: ref41 doi: 10.1109/5.726791 – year: 2018 ident: ref40 article-title: Adversarial logit pairing – ident: ref7 doi: 10.4204/EPTCS.257.3 – year: 2015 ident: ref16 article-title: Adversarial autoencoders – ident: ref31 doi: 10.1109/CVPR.2017.645 – start-page: 5967 year: 2017 ident: ref30 article-title: Fader networks: Manipulating images by sliding attributes publication-title: Proc Int Conf Neural Inf Process – ident: ref21 doi: 10.24963/ijcai.2018/543 – ident: ref11 doi: 10.1109/SP.2016.41 – ident: ref12 doi: 10.1109/SP.2017.49 – start-page: 2642 year: 2017 ident: ref17 article-title: Conditional image synthesis with auxiliary classifier GANs publication-title: Proc Int Conf Mach Learn – start-page: 2672 year: 2014 ident: ref4 article-title: Generative adversarial nets publication-title: Proc Int Conf Neural Inf Process – year: 2017 ident: ref20 article-title: Generative adversarial trainer: Defense to adversarial perturbations with GAN – start-page: 2687 year: 2018 ident: ref19 article-title: Learning to Attack: Adversarial Transformation Networks publication-title: Proc 32rd AAAI Conf Artif Intell – ident: ref26 doi: 10.1201/9781351251389-8 – start-page: 10 year: 2018 ident: ref13 article-title: EAD: Elastic-net attacks to deep neural networks via adversarial examples publication-title: Proc 32nd AAAI Conf Artif Intell – year: 2015 ident: ref34 article-title: Adam: A method for stochastic optimization publication-title: Proc Int Conf Learn Representations – year: 2017 ident: ref10 article-title: A theoretical framework for robustness of (deep) classifiers against adversarial examples publication-title: Proc Int Conf Learn Representations Workshop – year: 2014 ident: ref5 article-title: Auto-encoding variational Bayes publication-title: Proc Int Conf Learn Representations – start-page: 52 year: 2018 ident: ref22 article-title: IntroVAE: Introspective variational autoencoders for photographic image synthesis publication-title: Proc Int Conf Neural Inf Process – ident: ref25 doi: 10.1109/CVPR.2016.282 – ident: ref18 doi: 10.1109/CVPR.2019.00453 – year: 2018 ident: ref42 article-title: Ensemble adversarial training: Attacks and defenses publication-title: Proc Int Conf Learn Representations – ident: ref36 doi: 10.1109/ICCV.2015.425 – ident: ref39 doi: 10.14722/ndss.2018.23198 – year: 2018 ident: ref14 article-title: Towards deep learning models resistant to adversarial attacks publication-title: Proc Int Conf Learn Representations – year: 2007 ident: ref38 article-title: Labeled faces in the wild: A database for studying face recognition in unconstrained environments – ident: ref3 doi: 10.1109/CVPR.2017.660 – ident: ref1 doi: 10.1109/CVPR.2016.90 – year: 2016 ident: ref37 article-title: TensorFlow: Large-scale machine learning on heterogeneous distributed systems publication-title: arXiv 1603 04467 – ident: ref35 doi: 10.1109/5.726791 – year: 2017 ident: ref33 article-title: BEGAN: Boundary equilibrium generative adversarial networks publication-title: arXiv 1703 10717 – year: 2015 ident: ref6 article-title: Explaining and harnessing adversarial examples publication-title: Proc Int Conf Learn Representations – start-page: 3581 year: 2014 ident: ref28 article-title: Semi-supervised learning with deep generative models publication-title: Proc Int Conf Neural Inf Process – start-page: 8312 year: 2018 ident: ref15 article-title: Constructing unrestricted adversarial examples with generative models publication-title: Proc Int Conf Neural Inf Process – start-page: 6513 year: 2017 ident: ref29 article-title: Good semi-supervised learning that requires a bad GAN publication-title: Proc Int Conf Neural Inf Process – ident: ref32 doi: 10.1109/CVPR.2015.7298682 – year: 2014 ident: ref9 article-title: Intriguing properties of neural networks publication-title: Proc Int Conf Learn Representations – start-page: 1097 year: 2012 ident: ref2 article-title: ImageNet classification with deep convolutional neural networks publication-title: Proc Int Conf Neural Inf Process – year: 2018 ident: ref24 article-title: Adversarial spheres – ident: ref23 doi: 10.1109/CVPR.2015.7298640 – year: 2014 ident: ref27 article-title: Conditional generative adversarial nets publication-title: arXiv preprint arXiv 1411 1784
SSID	ssj0014503
Score	2.4915476
Snippet	Despite the great success of deep neural networks, the adversarial attack can cheat some well-trained classifiers by small permutations. In this paper, we...
SourceID	proquest pubmed crossref ieee
SourceType	Aggregation Database Index Database Enrichment Source Publisher
StartPage	1100
SubjectTerms	Adversarial attack Aerospace electronics Artificial neural networks Classifiers Face recognition Neural networks Permutations Sun supervised variational autoencoder Task analysis Toy manufacturing industry Training type I error
Title	Adversarial Attack Type I: Cheat Classifiers by Significant Changes
URI	https://ieeexplore.ieee.org/document/8807315 https://www.ncbi.nlm.nih.gov/pubmed/31442970 https://www.proquest.com/docview/2487437337 https://www.proquest.com/docview/2280552596
Volume	43
WOSCitedRecordID	wos000616309900025&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText	1
inHoldings	1
isFullTextHit
isPrint
journalDatabaseRights	– providerCode: PRVIEE databaseName: IEEE Electronic Library (IEL) customDbUrl: eissn: 2160-9292 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0014503 issn: 0162-8828 databaseCode: RIE dateStart: 19790101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3dS9xAEB-s9ME-aKtW4xdb6JtGL7ubbNa341AUWhFq4d7C7mZTjkpOvJzgf-_M5gMfbMG3kJ18sDOTmcl8_AC-ZyrRpZAqlpWVMdrbKjalFDGvRGl95rSrAmrJD3Vzk0-n-nYFToZeGO99KD7zp3QYcvnl3C3pV9kZypoS1FH-QSnV9moNGQOZBhRk9GBQwzGM6BtkRvrs7nb885qquPQpGrdMKALpExhJcE0Yxa_sUQBY-bevGWzO5cb73vYzrHe-JRu3wvAFVny9CRs9bgPr1HgTPr0aQrgFk4DJvDAkiWzcNMb9ZRSdsutzNqFPNQvAmbOKQLOZfWa_Zn9qKjBCnrC2OWGxDb8vL-4mV3EHrRA7kSZNrIyzVuuUO5mb1CZl5hwGyVxWqTOWa5f6zAiTc-HRvzMYauqscgl3lcpH2ljxFVbree13gWm0-aXGtZIK3hKb53gikEpJScsIkn6DC9fNHSf4i_sixB8jXQT-FMSfouNPBMfDNQ_t1I3_Um_R7g-U3cZHcNDzsegUc1FwDNBompNQEXwbllGlKE9iaj9fIg3PR2mKcWEWwU7L_-Hevdjsvf3MfVjjVPQSitQOYLV5XPpD-Oiemtni8QjldpofBbl9AdCr5D4
linkProvider	IEEE
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3da9RAEB9KFdQHq60f0bau4JumvexHku3bcbT06PUoeELfwu5mUw4lJ72c4H_vzOaDPthC30J28sHOTGYm8_ED-JJmiS6FzGJZWRmjva1iU0oR80qU1qdOuyqglsyy-Ty_vtZXW_Bt6IXx3ofiM39EhyGXX67chn6VHaOsZYI6yp8oKXnSdmsNOQOpAg4y-jCo4xhI9C0yI328uBpfTqmOSx-heUtFRjB9AmMJrgml-I5FChAr93ubweqc7TzufV_By867ZONWHF7Dlq93YadHbmCdIu_CiztjCPdgElCZ14ZkkY2bxrifjOJTNj1hE_pYswCduawINpvZv-z78qamEiPkCmvbE9Zv4MfZ6WJyHnfgCrETKmnizDhrtVbcydwom5Spcxgmc1kpZyzXTvnUCJNz4dHDMxhs6rRyCXdVlo-0seItbNer2r8HptHqlxrXSip5S2ye44lAKiWlLSNI-g0uXDd5nAAwfhUhAhnpIvCnIP4UHX8i-Dpc87udu_Eg9R7t_kDZbXwE-z0fi0411wXHEI3mOYksgs_DMioVZUpM7VcbpOH5SCmMDNMI3rX8H-7di82H_z_zEzw7X1zOitl0fvERnnMqgQkla_uw3dxu_AE8dX-a5fr2MEjvP3jt5p0
openUrl	ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Adversarial+Attack+Type+I%3A+Cheat+Classifiers+by+Significant+Changes&rft.jtitle=IEEE+transactions+on+pattern+analysis+and+machine+intelligence&rft.au=Tang%2C+Sanli&rft.au=Huang%2C+Xiaolin&rft.au=Chen%2C+Mingjian&rft.au=Sun%2C+Chengjin&rft.date=2021-03-01&rft.issn=0162-8828&rft.eissn=2160-9292&rft.volume=43&rft.issue=3&rft.spage=1100&rft.epage=1109&rft_id=info:doi/10.1109%2FTPAMI.2019.2936378&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_TPAMI_2019_2936378
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0162-8828&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0162-8828&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0162-8828&client=summon