Adversarial Examples: Opportunities and Challenges
Open accessThe DOI matches its Crossref record; 26 of the 26 references it cites were found in a registry or an index.
Abstract
Deep neural networks (DNNs) have shown huge superiority over humans in image recognition, speech processing, autonomous vehicles, and medical diagnosis. However, recent studies indicate that DNNs are vulnerable to adversarial examples (AEs), which are designed by attackers to fool deep learning models. Different from real examples, AEs can mislead the model to predict incorrect outputs while hardly be distinguished by human eyes, therefore threaten security-critical deep-learning applications. In recent years, the generation and defense of AEs have become a research hotspot in the field of artificial intelligence (AI) security. This article reviews the latest research progress of AEs. First, we introduce the concept, cause, characteristics, and evaluation metrics of AEs, then give a survey on the state-of-the-art AE generation methods with the discussion of advantages and disadvantages. After that, we review the existing defenses and discuss their limitations. Finally, future research opportunities and challenges on AEs are prospected.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155
Image quality assessment: from error visibility to structural similarity
Wang, Z., Bovik, A. C., Sheikh, H. R., & Simoncelli, E. P. (2004) · IEEE Transactions on Image Processing · doi:10.1109/tip.2003.819861 · Cited by 57,892
Rethinking the Inception Architecture for Computer Vision
Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., & Wojna, Z. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.308 · Cited by 32,138
Explaining and Harnessing Adversarial Examples
Goodfellow, I. J., Shlens, J., & Szegedy, C. (2014) · arXiv · doi:10.48550/arxiv.1412.6572 · Cited by 10,381
Towards Evaluating the Robustness of Neural Networks
Carlini, N., & Wagner, D. (2017) · 2017 IEEE Symposium on Security and Privacy (SP) · doi:10.1109/sp.2017.49 · Cited by 8,859
Intriguing properties of neural networks
Szegedy, C., Zaremba, W., Sutskever, I., Bruna, J., Erhan, D., Goodfellow, I., & Fergus, R. (2013) · arXiv · doi:10.48550/arxiv.1312.6199 · Cited by 8,143
Adversarial examples in the physical world
Kurakin, A., Goodfellow, I. J., & Bengio, S. (2018) · Artificial Intelligence Safety and Security · doi:10.1201/9781351251389-8 · Cited by 5,375
DeepFool: A Simple and Accurate Method to Fool Deep Neural Networks
Moosavi-Dezfooli, S.-M., Fawzi, A., & Frossard, P. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.282 · Cited by 5,317 · cited 2×
“Mohsen et al. [34] proposed a non-targeted attack method based on L 2 distance norm, called DeepFool.”
BackgroundMethodThe Limitations of Deep Learning in Adversarial Settings
Papernot, N., McDaniel, P., Jha, S., Fredrikson, M., Celik, Z. B., & Swami, A. (2016) · 2016 IEEE European Symposium on Security and Privacy (EuroS&P) · doi:10.1109/eurosp.2016.36 · Cited by 3,993
Practical Black-Box Attacks against Machine Learning
Papernot, N., McDaniel, P., Goodfellow, I., Jha, S., Celik, Z. B., & Swami, A. (2017) · Proceedings of the 2017 ACM on Asia Conference on Computer and Communications Security · doi:10.1145/3052973.3053009 · Cited by 3,495 · cited 2×
“Therefore, an attacker can use AEs to attack the models that perform the same task, which means that an attacker can construct AEs in the known ML model and then attack related unknown models [29].”
BackgroundUniversal Adversarial Perturbations
Moosavi-Dezfooli, S.-M., Fawzi, A., Fawzi, O., & Frossard, P. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.17 · Cited by 2,680
One Pixel Attack for Fooling Deep Neural Networks
Su, J., Vargas, D. V., & Sakurai, K. (2019) · IEEE Transactions on Evolutionary Computation · doi:10.1109/tevc.2019.2890858 · Cited by 2,406 · cited 1×
“Su et al. [40] proposed an extreme one pixel attack method which only changes one pixel for each image to construct AEs.”
BackgroundAdversarial machine learning at scale
Kurakin, A., Goodfellow, I., & Bengio, S. (2016) · arXiv · doi:10.48550/arxiv.1611.01236 · Cited by 2,325
Ensemble Adversarial Training: Attacks and Defenses
Tramèr, F., Kurakin, A., Papernot, N., Goodfellow, I., Boneh, D., & McDaniel, P. (2017) · arXiv · doi:10.48550/arxiv.1705.07204 · Cited by 2,176 · cited 2×
“Tramr et al. [39] proposed an ensemble attack RAND+FGSM.”
BackgroundAdversarial Examples: Attacks and Defenses for Deep Learning
Yuan, X., He, P., Zhu, Q., & Li, X. (2019) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2018.2886017 · Cited by 1,733 · cited 1×
“Recently, in addition to typical AE construction methods introduced above, a lot of other attack methods have been proposed [48], [49].”
BackgroundDelving into Transferable Adversarial Examples and Black-box Attacks
Liu, Y., Chen, X., Liu, C., & Song, D. (2016) · arXiv · doi:10.48550/arxiv.1611.02770 · Cited by 1,310
The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation
Brundage, M., Avin, S., Clark, J., Toner, H., Eckersley, P., Garfinkel, B., Dafoe, A., Scharre, P., Zeitzoff, T., Filar, B., et al. (2018) · arXiv · doi:10.48550/arxiv.1802.07228 · Cited by 849 · cited 1×
“In order to mitigate poisoning attacks and evasion attacks, a lot of defenses have been proposed [10], [11], [21]–[23].”
BackgroundThe Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation
Brundage, M., Avin, S., Clark, J., Toner, H., Eckersley, P., Garfinkel, B., Dafoe, A., Scharre, P., Zeitzoff, T., Filar, B., et al. (2018) · Apollo - University of Cambridge Repository · doi:10.17863/cam.22520 · Cited by 849
Lowd, D., & Meek, C. (2005) · Proceedings of the eleventh ACM SIGKDD international conference on Knowledge discovery in data mining · doi:10.1145/1081870.1081950 · Cited by 755 · cited 1×
“Lowd and Meek [19] proposed the concept of adversarial learning, in which an adversary conducts an attack to minimize a cost function.”
BackgroundDetecting Adversarial Samples from Artifacts
Feinman, R., Curtin, R. R., Shintre, S., & Gardner, A. B. (2017) · arXiv · doi:10.48550/arxiv.1703.00410 · Cited by 606
Brown, T. B., Mané, D., Roy, A., Abadi, M., & Gilmer, J. (2017) · arXiv · doi:10.48550/arxiv.1712.09665 · Cited by 509 · cited 1×
“ound of the minimum adversarial perturbation required to misclassify the image. CLEVER is the first attack-independent method and can evaluate the intrinsic robustness of neural networks. Brown et al. [55] proposed adversarial patch which does not need to subtly transform an existing image into another and can be place…”
MethodPreparing for the future of Artificial Intelligence
Bundy, A. (2017) · AI & SOCIETY · doi:10.1007/s00146-016-0685-0 · Cited by 425
A Survey on Security Threats and Defensive Techniques of Machine Learning: A Data Driven View
Liu, Q., Li, P., Zhao, W., Cai, W., Yu, S., & Leung, V. C. M. (2018) · IEEE Access · doi:10.1109/access.2018.2805680 · Cited by 421
ANTIDOTE: understanding and defending against poisoning of anomaly detectors
Rubinstein, B. I. P., Nelson, B., Huang, L., Joseph, A. D., Lau, S.-H., Rao, S., Taft, N., & Tygar, J. D. (2009) · Proceedings of the 9th ACM SIGCOMM conference on Internet measurement · doi:10.1145/1644893.1644895 · Cited by 331 · cited 1×
“In order to mitigate poisoning attacks and evasion attacks, a lot of defenses have been proposed [10], [11], [21]–[23].”
BackgroundA general framework for adversarial examples with objectives
Sharif, M., Bhagavatula, S., Bauer, L., & Reiter, M. K. (2019) · ACM Transactions on Privacy and Security · doi:10.1145/3317611 · Cited by 200 · cited 1×
“AEs bring a great threat to the security-critical AI applications such as face payment [64], medical systems [65], and autonomous vehicles [66], [67] based on image recognition in deep learning.”
Background
Cited by
Fed-IIoT: A Robust Federated Malware Detection Architecture in Industrial IoT
Taheri, R., Shojafar, M., Alazab, M., & Tafazolli, R. (2021) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2020.3043458 · Cited by 218
Artificial Intelligence Security: Threats and Countermeasures
Hu, Y., Kuang, W., Qin, Z., Li, K., Zhang, J., Gao, Y., Li, W., & Li, K. (2023) · ACM Computing Surveys · doi:10.1145/3487890 · Cited by 196
“Grey-box attacks are not common in practice [174].”
BackgroundSelf-citationAlotaibi, A., & Rassam, M. A. (2023) · Future Internet · doi:10.3390/fi15020062 · Cited by 176
“Never-theless, most defense techniques are ineffective against black-box attacks [95] or need more experimentation, as in [44].”
Challenges and Countermeasures for Adversarial Attacks on Deep Reinforcement Learning
Ilahi, I., Usama, M., Qadir, J., Janjua, M. U., Al-Fuqaha, A., Hoang, D. T., & Niyato, D. (2022) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2021.3111139 · Cited by 170
Privacy and Security of Big Data in AI Systems: A Research and Standards Perspective
Dilmaghani, S., Brust, M. R., Danoy, G., Cassagnes, N., Pecero, J., & Bouvry, P. (2019) · 2019 IEEE International Conference on Big Data (Big Data) · doi:10.1109/bigdata47090.2019.9006283 · Cited by 111
“The novel data is then fed into the trained model, named as Apply phase , where an adversary can penetrate the system and modify the results in his favor [11], [30], [31].”
BackgroundMethod
Metadata
- Authors
- Zhang, J., & Li, C.
- Venue
- IEEE Transactions on Neural Networks and Learning Systems
- Published
- 2019
- Volume
- 31
- Issue
- 7
- Pages
- 1-16
- ISSN
- 2162-237X (ISSN-L)
- 2162-2388
- OpenAlex
- W2891828758
- Cited by
- 236
- References
- 26
- FWCI
- 11.40
- Type
- journal article
- Language
- en
- Open access
- green
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.