NeuronInspect: Detecting Backdoors in Neural Networks via Output Explanations
arXiv:1911.07399
Abstract
Deep neural networks have achieved state-of-the-art performance on various tasks. However, lack of interpretability and transparency makes it easier for malicious attackers to inject trojan backdoor into the neural networks, which will make the model behave abnormally when a backdoor sample with a specific trigger is input. In this paper, we propose NeuronInspect, a framework to detect trojan backdoors in deep neural networks via output explanation techniques. NeuronInspect first identifies the existence of backdoor attack targets by generating the explanation heatmap of the output layer. We observe that generated heatmaps from clean and backdoored models have different characteristics. Therefore we extract features that measure the attributes of explanations from an attacked model namely: sparse, smooth and persistent. We combine these features and use outlier detection to figure out the outliers, which is the set of attack targets. We demonstrate the effectiveness and efficiency of NeuronInspect on MNIST digit recognition dataset and GTSRB traffic sign recognition dataset. We extensively evaluate NeuronInspect on different attack scenarios and prove better robustness and effectiveness over state-of-the-art trojan backdoor detection techniques Neural Cleanse by a great margin.
8 pages, 13 figures
References in corpus (2)
Cited by in corpus (6)
- Backdoor Attacks and Countermeasures on Deep Learning: A Comprehensive Review
- An Embarrassingly Simple Approach for Trojan Attack in Deep Neural Networks
- Be Careful about Poisoned Word Embeddings: Exploring the Vulnerability of the Embedding Layers in NLP Models
- Poisoning Deep Reinforcement Learning Agents with In-Distribution Triggers
- EX-RAY: Distinguishing Injected Backdoor from Natural Features in Neural Networks by Examining Differential Feature Symmetry
- Cassandra: Detecting Trojaned Networks from Adversarial Perturbations