DeepBaR: Fault Backdoor Attack on Deep Neural Network Layers
Not on the map. Why?
Abstract (arXiv)
Machine Learning using neural networks has received prominent attention recently because of its success in solving a wide variety of computational tasks, in particular in the field of computer vision. However, several works have drawn attention to potential security risks involved with the training and implementation of such networks. In this work, we introduce DeepBaR, a novel approach that implants backdoors on neural networks by faulting their behavior at training, especially during fine-tuning. Our technique aims to generate adversarial samples by optimizing a custom loss function that mimics the implanted backdoors while adding an almost non-visible trigger in the image. We attack three popular convolutional neural network architectures and show that DeepBaR attacks have a success rate of up to 98.30%. Furthermore, DeepBaR does not significantly affect the accuracy of the attacked networks after deployment when non-malicious inputs are given. Remarkably, DeepBaR allows attackers to choose an input that looks similar to a given class, from a human perspective, but that will be classified as belonging to an arbitrary target class.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 33 of its 68 references that are in the map. First posted on arXiv 30 Jul 2024.
Builds on
33 of 68 references.
References · 33
- 2015Deep Residual Learning for Image Recognitioncited by 51,809
- 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
- 2016Densely Connected Convolutional Networkscited by 6,691
- 2017BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chaincited by 1,052
- 2018Black-box Adversarial Attacks with Limited Queries and Informationcited by 499
- 2020Backdoor Learning: A Surveycited by 412
- 2020Reflection Backdoor: A Natural Backdoor Attack on Deep Neural Networkscited by 341
- 2021Anti-Backdoor Learning: Training Clean Models on Poisoned Datacited by 258
- 2019Universal Litmus Patterns: Revealing Backdoor Attacks in CNNscited by 143
- 2020RAB: Provable Robustness Against Backdoor Attackscited by 115
- 2019TBT: Targeted Neural Network Attack with Bit Trojancited by 112
- 2019Bit-Flip Attack: Crushing Neural Network with Progressive Bit Searchcited by 106
21 more
- 2021CRFL: Certifiably Robust Federated Learning against Backdoor Attackscited by 91
- 2021Backdoor Attack in the Physical Worldcited by 82
- 2020On Certifying Robustness against Backdoor Attacks via Randomized Smoothingcited by 74
- 2020An Analysis of Adversarial Attacks and Defenses on Autonomous Driving Modelscited by 67
- 2018Backdooring Convolutional Neural Networks via Targeted Weight Perturbationscited by 66
- 2019Terminal Brain Damage: Exposing the Graceless Degradation in Deep Neural Networks Under Hardware Fault Attackscited by 55
- 2020Can Adversarial Weight Perturbations Inject Neural Backdoors?cited by 51
- 2021Targeted Attack against Deep Neural Networks via Flipping Limited Weight Bitscited by 49
- 2020Computing Systems for Autonomous Driving: State-of-the-Art and Challengescited by 47
- 2021DeepSteal: Advanced Model Extractions Leveraging Efficient Weight Stealing in Memoriescited by 42
- 2020SNIFF: Reverse Engineering of Neural Networks with Fault Attackscited by 22
- 2019RowHammer: A Retrospectivecited by 19
- 2021RADAR: Run-time Adversarial Weight Attack Detection and Accuracy Recoverycited by 18
- 2021RA-BNN: Constructing Robust & Accurate Binary Neural Network to Simultaneously Defend Adversarial Bit-Flip Attack and Improve Accuracycited by 11
- 2018PoTrojan: powerful neural-level trojan designs in deep learning modelscited by 9
- 2021HASHTAG: Hash Signatures for Online Detection of Fault-Injection Attacks on Deep Neural Networkscited by 9
- 2021FooBaR: Fault Fooling Backdoor Attack on Neural Network Trainingcited by 6
- 2023Retrospective: Flipping Bits in Memory Without Accessing Them: An Experimental Study of DRAM Disturbance Errorscited by 5
- 2023Comparative Evaluation of Recent Universal Adversarial Perturbations in Image Classificationcited by 4
- 2024DeepNcode: Encoding-Based Protection against Bit-Flip Attacks on Neural Networkscited by 3
- 2014Formal verification of a software countermeasure against instruction skip attackscited by 2
Built on by
Citing papers per year: 2026: 1.
1 paper.