Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
fatcat! About Guide Changelog
Login/Signup

Attention-Guided Digital Adversarial Patches on Visual Detection release_rev_1cef1b70-b760-4fe7-94ba-2e98e270cdd9

by Dapeng Lang, Deyun Chen, Ran Shi, Yongjun He

Entity Metadata (schema)

abstracts[] {'sha1': 'ed43aa3c22ad8c0c767f62088e050dded629147c', 'content': 'Deep learning has been widely used in the field of image classification and image recognition and achieved positive practical results. However, in recent years, a number of studies have found that the accuracy of deep learning model based on classification greatly drops when making only subtle changes to the original examples, thus realizing the attack on the deep learning model. The main methods are as follows: adjust the pixels of attack examples invisible to human eyes and induce deep learning model to make the wrong classification; by adding an adversarial patch on the detection target, guide and deceive the classification model to make it misclassification. Therefore, these methods have strong randomness and are of very limited use in practical application. Different from the previous perturbation to traffic signs, our paper proposes a method that is able to successfully hide and misclassify vehicles in complex contexts. This method takes into account the complex real scenarios and can perturb with the pictures taken by a camera and mobile phone so that the detector based on deep learning model cannot detect the vehicle or misclassification. In order to improve the robustness, the position and size of the adversarial patch are adjusted according to different detection models by introducing the attachment mechanism. Through the test of different detectors, the patch generated in the single target detection algorithm can also attack other detectors and do well in transferability. Based on the experimental part of this paper, the proposed algorithm is able to significantly lower the accuracy of the detector. Affected by the real world, such as distance, light, angles, resolution, etc., the false classification of the target is realized by reducing the confidence level and background of the target, which greatly perturbs the detection results of the target detector. In COCO Dataset 2017, it reveals that the success rate of this algorithm reaches 88.7%.', 'mimetype': 'application/xml+jats', 'lang': None}
container {'state': 'active', 'ident': 'sdme5pnua5auzcsjgqmqefb66m', 'revision': '9e0b14b0-003d-4e72-9bdd-22597ceef37c', 'redirect': None, 'extra': {'country': 'gb', 'default_license': 'CC-BY', 'doaj': {'archive': ['LOCKSS', 'Portico'], 'as_of': '2022-07-06', 'default_license': 'CC-BY', 'seal': True}, 'kbart': {'clockss': {'year_spans': [[2008, 2016]]}, 'lockss': {'year_spans': [[2017, 2018]]}, 'portico': {'year_spans': [[2011, 2022]]}, 'scholarsportal': {'year_spans': [[2008, 2015]]}}, 'languages': ['en'], 'publisher_type': 'big5', 'sherpa_romeo': {'color': 'green'}, 'urls': ['https://www.hindawi.com/journals/scn/', 'http://www.interscience.wiley.com/journal/security', 'http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1939-0122']}, 'edit_extra': None, 'name': 'Security and Communication Networks', 'container_type': None, 'publication_status': None, 'publisher': 'Hindawi Limited', 'issnl': '1939-0122', 'issne': '1939-0122', 'issnp': '1939-0114', 'wikidata_qid': 'Q15762718'}
container_id sdme5pnua5auzcsjgqmqefb66m
contribs[] {'index': 0, 'creator_id': 'vhv5jvp3gbbobelpgpynsgbfd4', 'creator': None, 'raw_name': 'Dapeng Lang', 'given_name': 'Dapeng', 'surname': 'Lang', 'role': 'author', 'raw_affiliation': 'School of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150001, China', 'extra': {'seq': 'first'}}
{'index': 1, 'creator_id': 'twnxbqzhszcgdb23mituxupo5e', 'creator': None, 'raw_name': 'Deyun Chen', 'given_name': 'Deyun', 'surname': 'Chen', 'role': 'author', 'raw_affiliation': 'School of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150001, China', 'extra': None}
{'index': 2, 'creator_id': None, 'creator': None, 'raw_name': 'Ran Shi', 'given_name': 'Ran', 'surname': 'Shi', 'role': 'author', 'raw_affiliation': 'School of Computer Science and Technology, Heilongjiang University, Harbin 150001, China', 'extra': None}
{'index': 3, 'creator_id': 'a3owjst5mbaxdlyloud4dl62ze', 'creator': None, 'raw_name': 'Yongjun He', 'given_name': 'Yongjun', 'surname': 'He', 'role': 'author', 'raw_affiliation': 'School of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150001, China', 'extra': None}
{'index': None, 'creator_id': None, 'creator': None, 'raw_name': 'Tom Chen', 'given_name': 'Tom', 'surname': 'Chen', 'role': 'editor', 'raw_affiliation': None, 'extra': None}
ext_ids {'doi': '10.1155/2021/6637936', 'wikidata_qid': None, 'isbn13': None, 'pmid': None, 'pmcid': None, 'core': None, 'arxiv': None, 'jstor': None, 'ark': None, 'mag': None, 'doaj': None, 'dblp': None, 'oai': None, 'hdl': None}
files
filesets
issue
language en
license_slug CC-BY
number
original_title
pages 1-11
publisher Hindawi Limited
refs[] {'index': 0, 'target_release_id': None, 'extra': {'doi': '10.1016/j.ins.2020.08.087'}, 'key': '1', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 1, 'target_release_id': None, 'extra': {'authors': ['C. Chen']}, 'key': '2', 'year': 2019, 'container_name': 'Advances in Neural Information Processing System', 'title': 'This looks like that: deep learning for interpretable image recognition', 'locator': None}
{'index': 2, 'target_release_id': None, 'extra': {'authors': ['J. Choi']}, 'key': '3', 'year': None, 'container_name': None, 'title': 'Gaussian YOLOv3: an accurate and fast object detector using localization uncertainty for autonomous driving', 'locator': '502'}
{'index': 3, 'target_release_id': None, 'extra': {'doi': '10.1109/iccv.2017.324'}, 'key': '4', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 4, 'target_release_id': None, 'extra': {'doi': '10.1155/2020/5890820'}, 'key': '5', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 5, 'target_release_id': None, 'extra': {'doi': '10.1109/cvpr.2017.690'}, 'key': '6', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 6, 'target_release_id': None, 'extra': {'doi': '10.1109/tpami.2016.2577031'}, 'key': '7', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 7, 'target_release_id': None, 'extra': {'doi': '10.1109/access.2020.2964424'}, 'key': '8', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 8, 'target_release_id': None, 'extra': {'doi': '10.1109/taslp.2016.2615239'}, 'key': '9', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 9, 'target_release_id': None, 'extra': {'authors': ['Y. Fu']}, 'key': '10', 'year': 2019, 'container_name': None, 'title': 'Self-similarity grouping: a simple unsupervised cross domain adaptation approach for person re-identification', 'locator': None}
{'index': 10, 'target_release_id': None, 'extra': {'authors': ['K. Eykholt']}, 'key': '11', 'year': 2018, 'container_name': None, 'title': 'Physical adversarial examples for object detectors', 'locator': None}
{'index': 11, 'target_release_id': None, 'extra': {'authors': ['L. Yujia'], 'volume': '2017'}, 'key': '12', 'year': 2017, 'container_name': 'Security and Communication Networks', 'title': 'Protecting privacy in shared photos via adversarial examples based stealth', 'locator': None}
{'index': 12, 'target_release_id': None, 'extra': {'doi': '10.1109/cvprw.2018.00210'}, 'key': '13', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 13, 'target_release_id': None, 'extra': {'doi': '10.1109/tnnls.2018.2886017'}, 'key': '14', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 14, 'target_release_id': None, 'extra': {'authors': ['D. Lang'], 'volume': '10'}, 'key': '15', 'year': 2014, 'container_name': 'Journal of Computational Information Systems', 'title': 'A state space abstract algorithm of incremental data recognition based on model checking', 'locator': '1731'}
{'index': 15, 'target_release_id': None, 'extra': {'authors': ['H. Chang'], 'doi': '10.1109/cvpr.2018.00012'}, 'key': '16', 'year': None, 'container_name': None, 'title': 'PairedCycleGAN: asymmetric style transfer for applying and removing makeup', 'locator': None}
{'index': 16, 'target_release_id': None, 'extra': {'authors': ['J. Marniemi']}, 'key': '17', 'year': 1975, 'container_name': None, 'title': 'NO need to worry about adversarial examples in object detection in autonomous vehicles', 'locator': None}
{'index': 17, 'target_release_id': None, 'extra': {'doi': '10.1016/j.ins2020.12.042'}, 'key': '18', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 18, 'target_release_id': None, 'extra': {'doi': '10.1109/tcsvt.2019.2944430'}, 'key': '19', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 19, 'target_release_id': None, 'extra': {'authors': ['P. Isola']}, 'key': '20', 'year': None, 'container_name': None, 'title': 'Image-to-Image translation with conditional adversarial networks', 'locator': '5967'}
{'index': 20, 'target_release_id': None, 'extra': {'doi': '10.1007/s10845-020-01639-1'}, 'key': '21', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 21, 'target_release_id': None, 'extra': {'authors': ['L. Jiang']}, 'key': '22', 'year': 2020, 'container_name': None, 'title': 'Cycle-consistent adversarial GAN', 'locator': None}
{'index': 22, 'target_release_id': None, 'extra': {'authors': ['A. Chattopadhay']}, 'key': '23', 'year': None, 'container_name': None, 'title': 'Grad-CAM++: generalized gradient-based visual explanations for deep convolutional networks', 'locator': '839'}
{'index': 23, 'target_release_id': None, 'extra': {'doi': '10.1007/s11263-019-01228-7'}, 'key': '24', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 24, 'target_release_id': None, 'extra': {'doi': '10.1002/jcd.21358'}, 'key': '25', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 25, 'target_release_id': None, 'extra': {'authors': ['K. Eshghi'], 'doi': '10.1109/icde.2016.7498284'}, 'key': '26', 'year': None, 'container_name': None, 'title': 'The CRO kernel: using concomitant rank order hashes for sparse high dimensional randomized feature maps', 'locator': None}
{'index': 26, 'target_release_id': None, 'extra': {'doi': '10.1016/j.asoc.2020.106939'}, 'key': '27', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 27, 'target_release_id': None, 'extra': {'doi': '10.1609/aaai.v33i01.33011028'}, 'key': '28', 'year': None, 'container_name': None, 'title': None, 'locator': None}
{'index': 28, 'target_release_id': None, 'extra': {'doi': '10.1109/iccv.2019.00055'}, 'key': '29', 'year': None, 'container_name': None, 'title': None, 'locator': None}
release_date 2021-04-07
release_stage published
release_type article-journal
release_year 2021
subtitle
title Attention-Guided Digital Adversarial Patches on Visual Detection
version
volume 2021
webcaptures
withdrawn_date
withdrawn_status
withdrawn_year
work_id fcyhsmu66fcs3ovp25ri53la5y

Extra Metadata (raw JSON)

crossref.alternative-id ['6637936', '6637936']
crossref.funder [{'award': ['KY10600200021'], 'name': '2020 Information Security Software Project, Network Threat Depth Analysis Software'}]
crossref.license [{'URL': 'https://creativecommons.org/licenses/by/4.0/', 'content-version': 'unspecified', 'delay-in-days': 0, 'start': '2021-04-07T00:00:00Z'}]
crossref.subject ['Computer Networks and Communications', 'Information Systems']
crossref.type journal-article