Model-Contrastive Learning for Backdoor Defense
Not on the map. Why?
Abstract (arXiv)
Due to the popularity of Artificial Intelligence (AI) techniques, we are witnessing an increasing number of backdoor injection attacks that are designed to maliciously threaten Deep Neural Networks (DNNs) causing misclassification. Although there exist various defense methods that can effectively erase backdoors from DNNs, they greatly suffer from both high Attack Success Rate (ASR) and a non-negligible loss in Benign Accuracy (BA). Inspired by the observation that a backdoored DNN tends to form a new cluster in its feature spaces for poisoned data, in this paper we propose a novel two-stage backdoor defense method, named MCLDef, based on Model-Contrastive Learning (MCL). In the first stage, our approach performs trigger inversion based on trigger synthesis, where the resultant trigger can be used to generate poisoned data. In the second stage, under the guidance of MCL and our defined positive and negative pairs, MCLDef can purify the backdoored model by pulling the feature representations of poisoned data towards those of their clean data counterparts. Due to the shrunken cluster of poisoned data, the backdoor formed by end-to-end supervised learning is eliminated. Comprehensive experimental results show that, with only 5% of clean data, MCLDef significantly outperforms state-of-the-art defense methods by up to 95.79% reduction in ASR, while in most cases the BA degradation can be controlled within less than 2%. Our code is available at https://github.com/WeCanShow/MCL.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 33 of its 44 references that are in the map. First posted on arXiv 9 May 2022.
Builds on
33 of 44 references.
References · 33
- 2015Deep Residual Learning for Image Recognitioncited by 51,809
- 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
- 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
- 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
- 2016Wide Residual Networkscited by 3,057
- 2016Towards Evaluating the Robustness of Neural Networkscited by 2,976
- 2020Exploring Simple Siamese Representation Learningcited by 2,853
- 2020Big Self-Supervised Models are Strong Semi-Supervised Learnerscited by 1,375
- 2017Mixed Precision Trainingcited by 1,135
- 2017Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoningcited by 1,012
- 2017ChestX-ray8: Hospital-scale Chest X-ray Database and Benchmarks on Weakly-Supervised Classification and Localization of Common Thorax Diseasescited by 969
- 2021Model-Contrastive Federated Learningcited by 762
21 more
- 2019Self-supervised Visual Feature Learning with Deep Neural Networks: A Surveycited by 656
- 2018Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networkscited by 550
- 2020Backdoor Learning: A Surveycited by 412
- 2020Reflection Backdoor: A Natural Backdoor Attack on Deep Neural Networkscited by 341
- 2021MagFace: A Universal Representation for Face Recognition and Quality Assessmentcited by 332
- 2021Neural Attention Distillation: Erasing Backdoor Triggers from Deep Neural Networkscited by 313
- 2019Label-Consistent Backdoor Attackscited by 285
- 2021Anti-Backdoor Learning: Training Clean Models on Poisoned Datacited by 258
- 2019A new Backdoor Attack in CNNs by training set corruption without label poisoningcited by 237
- 2021Adversarial Unlearning of Backdoors via Implicit Hypergradientcited by 160
- 2021Rethinking the Backdoor Attacks' Triggers: A Frequency Perspectivecited by 155
- 2022Backdoor Defense via Decoupling the Training Processcited by 149
- 2020Pillar-based Object Detection for Autonomous Drivingcited by 137
- 2018Backdoor Embedding in Convolutional Neural Network Models via Invisible Perturbationcited by 131
- 2019Demon in the Variant: Statistical Analysis of DNNs for Robust Backdoor Contamination Detectioncited by 121
- 2020Strong Data Augmentation Sanitizes Poisoning and Backdoor Attacks Without an Accuracy Tradeoffcited by 86
- 2019Defending Neural Backdoors via Generative Distribution Modelingcited by 82
- 2021Black-box Detection of Backdoor Attacks with Limited Information and Datacited by 71
- 2021Topological Detection of Trojaned Neural Networkscited by 34
- 2022Post-Training Detection of Backdoor Attacks for Two-Class and Multi-Attack Scenarioscited by 29
- 2020FDA3 : Federated Defense Against Adversarial Attacks for Cloud-Based IIoT Applicationscited by 10
Built on by
Citing papers per year: 2022: 1.
1 paper.