Edge Detectors Can Make Deep Convolutional Neural Networks More Robust
Not on the map. Why?
Abstract (arXiv)
Deep convolutional neural networks (DCNN for short) are vulnerable to examples with small perturbations. Improving DCNN's robustness is of great significance to the safety-critical applications, such as autonomous driving and industry automation. Inspired by the principal way that human eyes recognize objects, i.e., largely relying on the shape features, this paper first employs the edge detectors as layer kernels and designs a binary edge feature branch (BEFB for short) to learn the binary edge features, which can be easily integrated into any popular backbone. The four edge detectors can learn the horizontal, vertical, positive diagonal, and negative diagonal edge features, respectively, and the branch is stacked by multiple Sobel layers (using edge detectors as kernels) and one threshold layer. The binary edge features learned by the branch, concatenated with the texture features learned by the backbone, are fed into the fully connected layers for classification. We integrate the proposed branch into VGG16 and ResNet34, respectively, and conduct experiments on multiple datasets. Experimental results demonstrate the BEFB is lightweight and has no side effects on training. And the accuracy of the BEFB integrated models is better than the original ones on all datasets when facing FGSM, PGD, and C&W attacks. Besides, BEFB integrated models equipped with the robustness enhancing techniques can achieve better classification accuracy compared to the original models. The work in this paper for the first time shows it is feasible to enhance the robustness of DCNNs through combining both shape-like features and texture features.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 27 of its 40 references that are in the map. First posted on arXiv 26 Feb 2024.
Builds on
27 of 40 references.
References · 27
- 2015Deep Residual Learning for Image Recognitioncited by 51,809
- 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
- 2014Explaining and Harnessing Adversarial Examplescited by 7,287
- 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
- 2013Intriguing properties of neural networkscited by 5,114
- 2016Towards Evaluating the Robustness of Neural Networkscited by 2,976
- 2013Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computationcited by 2,031
- 2015DeepFool: a simple and accurate method to fool deep neural networkscited by 1,315
- 2018Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examplescited by 1,259
- 2016Adversarial Machine Learning at Scalecited by 934
- 2018Robustness May Be at Odds with Accuracycited by 873
- 2020Fast is better than free: Revisiting adversarial trainingcited by 723
15 more
- 2021Recent Advances in Adversarial Training for Adversarial Robustnesscited by 276
- 2018Adversarial Logit Pairingcited by 224
- 2018Adversarial Examples: Opportunities and Challengescited by 62
- 2021Subspace Adversarial Trainingcited by 41
- 2019Why ResNet Works? Residuals Generalizecited by 39
- 2022Gradient Obfuscation Gives a False Sense of Security in Federated Learningcited by 37
- 2020Adversarial Robustness of Stabilized NeuralODEs Might be from Obfuscated Gradientscited by 22
- 2022Recognizing Object by Components with Human Prior Knowledge Enhances Adversarial Robustness of Deep Neural Networkscited by 18
- 2023Injecting Logical Constraints into Neural Networks via Straight-Through Estimatorscited by 12
- 2022Part-Based Models Improve Adversarial Robustnesscited by 11
- 2022A2: Efficient Automated Attacker for Boosting Adversarial Trainingcited by 10
- 2022Are Straight-Through gradients and Soft-Thresholding all you need for Sparse Training?cited by 9
- 2021Recent Advances in Large Margin Learningcited by 7
- 2022Revisiting Outer Optimization in Adversarial Trainingcited by 4
- 2019Adversarial Margin Maximization Networkscited by 3
Built on by
Citing papers per year: 2024: 1; 2025: 1.
2 papers.