Latent Backdoor Attacks on Deep Neural Networks
Open accessThe DOI matches its Crossref record; 45 of the 45 references it cites were found in a registry or an index.
Abstract
Recent work proposed the concept of backdoor attacks on deep neural networks (DNNs), where misclassification rules are hidden inside normal models, only to be triggered by very specific inputs. However, these "traditional" backdoors assume a context where users train their own models from scratch, which rarely occurs in practice. Instead, users typically customize "Teacher" models already pretrained by providers like Google, through a process called transfer learning. This customization process introduces significant changes to models and disrupts hidden backdoors, greatly reducing the actual impact of backdoors in practice. In this paper, we describe latent backdoors, a more powerful and stealthy variant of backdoor attacks that functions under transfer learning. Latent backdoors are incomplete backdoors embedded into a "Teacher" model, and automatically inherited by multiple "Student" models through transfer learning. If any Student models include the label targeted by the backdoor, then its customization process completes the backdoor and makes it active. We show that latent backdoors can be quite effective in a variety of application contexts, and validate its practicality through real-world attacks against traffic sign recognition, iris identification of volunteers, and facial recognition of public figures (politicians). Finally, we evaluate 4 potential defenses, and find that only one is effective in disrupting latent backdoors, but might incur a cost in classification accuracy as tradeoff.
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
You Only Look Once: Unified, Real-Time Object Detection
Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.91 · Cited by 53,803
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
Parkhi, O. M., Vedaldi, A., & Zisserman, A. (2015) · Procedings of the British Machine Vision Conference 2015 · doi:10.5244/c.29.41 · Cited by 5,479
CNN Features Off-the-Shelf: An Astounding Baseline for Recognition
Razavian, A. S., Azizpour, H., Sullivan, J., & Carlsson, S. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops · doi:10.1109/cvprw.2014.131 · Cited by 4,451
The 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
Distillation as a Defense to Adversarial Perturbations Against Deep Neural Networks
Papernot, N., McDaniel, P., Wu, X., Jha, S., & Swami, A. (2016) · 2016 IEEE Symposium on Security and Privacy (SP) · doi:10.1109/sp.2016.41 · Cited by 2,982
Adversarial machine learning at scale
Kurakin, A., Goodfellow, I., & Bengio, S. (2016) · arXiv · doi:10.48550/arxiv.1611.01236 · Cited by 2,325
Robust physical-world attacks on deep learning visual classification
Eykholt, K., Evtimov, I., Fernandes, E., Li, B., Rahmati, A., Xiao, C., Prakash, A., Kohno, T., & Song, D. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00175 · Cited by 2,173
Obfuscated gradients give a false sense of security: Circumventing defenses to adversarial examples
Athalye, A., Carlini, N., & Wagner, D. (2018) · arXiv · doi:10.48550/arxiv.1802.00420 · Cited by 2,172 · cited 1×
“A number of defenses have been proposed [23, 32, 33, 38, 52], yet many have shown to be less effective against an adaptive attacker [4, 7, 8, 19].”
BackgroundFeature Squeezing: Detecting Adversarial Examples in Deep Neural Networks
Xu, W., Evans, D., & Qi, Y. (2018) · Proceedings 2018 Network and Distributed System Security Symposium · doi:10.14722/ndss.2018.23198 · Cited by 1,768
Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot Translation
Johnson, M., Schuster, M., Le, Q. V., Krikun, M., Wu, Y., Chen, Z., Thorat, N., Viégas, F., Wattenberg, M., Corrado, G., et al. (2017) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00065 · Cited by 1,747
Neural Cleanse: Identifying and Mitigating Backdoor Attacks in Neural Networks
Wang, B., Yao, Y., Shan, S., Li, H., Viswanath, B., Zheng, H., & Zhao, B. Y. (2019) · 2019 IEEE Symposium on Security and Privacy (SP) · doi:10.1109/sp.2019.00031 · Cited by 1,368
Self-citationExploiting Similarities among Languages for Machine Translation
Mikolov, T., Le, Q. V., & Sutskever, I. (2013) · arXiv (Cornell University) · doi:10.48550/arxiv.1309.4168 · Cited by 1,346
Delving 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
Trojaning Attack on Neural Networks
Liu, Y., Ma, S., Aafer, Y., Lee, W.-C., Zhai, J., Wang, W., & Zhang, X. (2018) · Proceedings 2018 Network and Distributed System Security Symposium · doi:10.14722/ndss.2018.23291 · Cited by 1,270
BadNets: Evaluating Backdooring Attacks on Deep Neural Networks
Gu, T., Liu, K., Dolan-Gavitt, B., & Garg, S. (2019) · IEEE Access · doi:10.1109/access.2019.2909068 · Cited by 1,238 · cited 8×
“Gu et al. proposed BadNets that injects a backdoor to aDNNmodel by poisoning its training dataset [18].”
BackgroundMethodInfluentialThe German Traffic Sign Recognition Benchmark: A multi-class classification competition
Stallkamp, J., Schlipsing, M., Salmen, J., & Igel, C. (2011) · The 2011 International Joint Conference on Neural Networks · doi:10.1109/ijcnn.2011.6033395 · Cited by 1,144
BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chain
Gu, T., Dolan-Gavitt, B., & Garg, S. (2017) · arXiv · doi:10.48550/arxiv.1708.06733 · Cited by 1,061
Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networks
Liu, K., Dolan-Gavitt, B., & Garg, S. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-00470-5_13 · Cited by 1,055
Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning
Chen, X., Liu, C., Li, B., Lu, K., & Song, D. (2017) · arXiv · doi:10.48550/arxiv.1712.05526 · Cited by 1,039 · cited 3×
“This fails against backdoor attacks [12], as injected samples do not affect the model’s performance on clean samples.”
BackgroundInfluentialMagNet: A Two-Pronged Defense against Adversarial Examples
Meng, D., & Chen, H. (2017) · Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/3133956.3134057 · Cited by 1,006
One-Shot Video Object Segmentation
Caelles, S., Maninis, K.-K., Pont-Tuset, J., Leal-Taixe, L., Cremers, D., & Van Gool, L. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.565 · Cited by 927 · cited 1×
“In a deep learning context, transfer learning has been shown to be effective in vision [5, 11, 42, 43], speech [13, 20, 24, 51], and text [22, 34].”
BackgroundManipulating Machine Learning: Poisoning Attacks and Countermeasures for Regression Learning
Jagielski, M., Oprea, A., Biggio, B., Liu, C., Nita-Rotaru, C., & Li, B. (2018) · 2018 IEEE Symposium on Security and Privacy (SP) · doi:10.1109/sp.2018.00057 · Cited by 861
Mogelmose, A., Trivedi, M. M., & Moeslund, T. B. (2012) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2012.2209421 · Cited by 684
Brown, T. B., Mané, D., Roy, A., Abadi, M., & Gilmer, J. (2017) · arXiv · doi:10.48550/arxiv.1712.09665 · Cited by 509
Detecting Backdoor Attacks on Deep Neural Networks by Activation Clustering
Chen, B., Carvalho, W., Baracaldo, N., Ludwig, H., Edwards, B., Lee, T., Molloy, I., & Srivastava, B. (2018) · arXiv · doi:10.48550/arxiv.1811.03728 · Cited by 485
Certified Defenses for Data Poisoning Attacks
Steinhardt, J., Koh, P. W., & Liang, P. (2017) · arXiv · doi:10.48550/arxiv.1706.03691 · Cited by 479
Robust Physical-World Attacks on Deep Learning Models
Eykholt, K., Evtimov, I., Fernandes, E., Li, B., Rahmati, A., Xiao, C., Prakash, A., Kohno, T., & Song, D. X. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1707.08945 · Cited by 444
Systematic Poisoning Attacks on and Defenses for Machine Learning in Healthcare
Mozaffari-Kermani, M., Sur-Kolay, S., Raghunathan, A., & Jha, N. K. (2015) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2014.2344095 · Cited by 383
Spectral Signatures in Backdoor Attacks
Tran, B., Li, J., & Mądry, A. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1811.00636 · Cited by 343
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×
“Defenses against poisoning attacks mostly focus on sanitizing training data and removing poisoned samples [6, 15, 21, 36, 44, 47].”
BackgroundMultilingual acoustic models using distributed deep neural networks
Heigold, G., Vanhoucke, V., Senior, A., Nguyen, P., Ranzato, M., Devin, M., & Dean, J. (2013) · 2013 IEEE International Conference on Acoustics, Speech and Signal Processing · doi:10.1109/icassp.2013.6639348 · Cited by 273
Liu, Y., Xie, Y., & Srivastava, A. (2017) · 2017 IEEE International Conference on Computer Design (ICCD) · doi:10.1109/iccd.2017.16 · Cited by 241
Poison Frogs! Targeted Clean-Label Poisoning Attacks on Neural Networks
Shafahi, A., Huang, W., Najibi, M., Suciu, O., Studer, C., Dumitraş, T., & Goldstein, T. (2018) · arXiv (Cornell University) · doi:10.3929/ethz-b-000460348 · Cited by 236
Casting out Demons: Sanitizing Training Data for Anomaly Sensors
Cretu, G. F., Stavrou, A., Locasto, M. E., Stolfo, S. J., & Keromytis, A. D. (2008) · 2008 IEEE Symposium on Security and Privacy (sp 2008) · doi:10.1109/sp.2008.11 · Cited by 226 · cited 1×
“Defenses against poisoning attacks mostly focus on sanitizing training data and removing poisoned samples [6, 15, 21, 36, 44, 47].”
BackgroundAdversarial Example Defense: Ensembles of Weak Defenses are not Strong
He, W., Wei, J., Chen, X., Carlini, N., & Song, D. X. (2017) · Cited by 221
Transfer learning for speech and language processing
Wang, D., & Zheng, T. F. (2015) · 2015 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA) · doi:10.1109/apsipa.2015.7415532 · Cited by 213
Transfer learning for Latin and Chinese characters with Deep Neural Networks
Ciresan, D. C., Meier, U., & Schmidhuber, J. (2012) · The 2012 International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2012.6252544 · Cited by 209
Pinto, N., Stone, Z., Zickler, T. E., & Cox, D. (2011) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2011.5981788 · Cited by 168
Transfer Learning for Speech Recognition on a Budget
Kunze, J., Kirsch, L., Kurenkov, I., Krug, A., Johannsmeier, J., & Stober, S. (2017) · Proceedings of the 2nd Workshop on Representation Learning for NLP · doi:10.18653/v1/w17-2620 · Cited by 141
With Great Training Comes Great Vulnerability: Practical Attacks against Transfer Learning
Wang, B., Yao, Y., Viswanath, B., Zheng, H.-T., & Zhao, B. Y. (2018) · USENIX Security Symposium · Cited by 139
Self-citationAn End-to-End System for Unconstrained Face Verification with Deep Convolutional Neural Networks
Chen, J.-C., Ranjan, R., Kumar, A., Chen, C., Patel, V. M., & Chellappa, R. (2015) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2015.55 · Cited by 99
Hu-Fu: Hardware and Software Collaborative Attack Framework Against Neural Networks
Li, W., Yu, J., Ning, X., Wang, P., Wei, Q., Wang, Y., & Yang, H. (2018) · 2018 IEEE Computer Society Annual Symposium on VLSI (ISVLSI) · doi:10.1109/isvlsi.2018.00093 · Cited by 58
Efficient Repair of Polluted Machine Learning Systems via Causal Unlearning
Cao, Y., Yu, A. F., Aday, A., Stahl, E., Merwine, J., & Yang, J. (2018) · ACM Asia Conference on Computer and Communications Security (ASIA CCS) · doi:10.1145/3196494.3196517 · Cited by 50
Neural TrojansPreprint
Liu, Y., Xie, Y., & Srivastava, A. (2017) · arXiv (Cornell University) · Cited by 29
Cited by
Li, Y., Jiang, Y., Li, Z., & Xia, S.-T. (2024) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2022.3182979 · Cited by 682
Flower: A Friendly Federated Learning Research Framework
Beutel, D. J., Topal, T., Mathur, A., Qiu, X., Fernandez-Marques, J., Gao, Y., Sani, L., Li, K. H., Parcollet, T., de Gusmão, P. P. B., et al. (2020) · arXiv · doi:10.48550/arxiv.2007.14390 · Cited by 622
“There has been tremendous progress is enabling the execution of deep learning models on mobile and embedded devices to infer user contexts and behaviors [1, 2, 3, 4, 5, 6, 7].”
BackgroundPrivacy and Robustness in Federated Learning: Attacks and Defenses
Lyu, L., Yu, H., Ma, X., Chen, C., Sun, L., Zhao, J., Yang, Q., & Yu, P. S. (2024) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2022.3216981 · Cited by 548
“…( e.g. , SIGN SGD [160]) or only sharing model predictions [158], [25], [87] in a black-box manner may result in more robust privacy protection in FL. Rethinking Current Privacy Attacks: There are several inherent weaknesses in current attacks that may limit their applicability in FL [161].”
BackgroundReflection Backdoor: A Natural Backdoor Attack on Deep Neural Networks
Liu, Y., Ma, X., Bailey, J., & Lu, F. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58607-2_11 · Cited by 531
When Machine Learning Meets Privacy: A Survey and Outlook
Liu, B., Ding, M., Shaham, S., Rahayu, W., Farokhi, F., & Lin, Z. (2022) · ACM Computing Surveys · doi:10.1145/3436755 · Cited by 455
Metadata
- Authors
- Yao, Y., Li, H., Zheng, H., & Zhao, B. Y.
- Venue
- Proceedings of the 2019 ACM SIGSAC Conference on Computer and Communications Security
- Publisher
- ACM
- Published
- 2019-11-06
- Pages
- 2041-2055
- OpenAlex
- W2985913519
- Cited by
- 343
- References
- 45
- FWCI
- 16.03
- Type
- conference paper
- Language
- en
- Open access
- hybrid
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.