Geometry-aware Instance-reweighted Adversarial Training

By Jingfeng Zhang, Jianing Zhu, Gang Niu, Bo Han, Masashi Sugiyama, Mohan Kankanhalli. International Conference on Learning Representations, 2020

318 citations recorded by the source corpus.

In adversarial machine learning, there was a common belief that robustness and accuracy hurt each other. The belief was challenged by recent studies where we can maintain the robustness and improve the accuracy. However, the other direction, whether we can keep the accuracy while improving the robustness, is conceptually and practically more interesting, since robust accuracy should be lower than standard accuracy for any model. In this paper, we show this direction is also promising. Firstly, we find even over-parameterized deep networks may still have insufficient model capacity, because adversarial training has an overwhelming smoothing effect. Secondly, given limited model capacity, we argue adversarial data should have unequal importance: geometrically speaking, a natural data point closer to/farther from the class boundary is less/more robust, and the corresponding adversarial data point should be assigned with larger/smaller weight. Finally, to implement the idea, we propose geometry-aware instance-reweighted adversarial training, where the weights are based on how difficult it is to attack a natural data point. Experiments show that our proposal boosts the robustness of standard adversarial training; combining two directions, we improve both robustness and accuracy of standard adversarial training.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/99a599d8fe56529f47e78243ed61250190f96196

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Yifan Zhang
2. Jian Liang
3. Jiashi Feng
4. Dapeng Hu
5. Bryan Hooi
6. Mingkui Tan
7. Trevor Darrell
8. Shuaicheng Niu

In Computer Vision and Pattern Recognition

1. Yifan Zhang
2. Jian Liang
3. Dapeng Hu
4. Jiashi Feng
5. Bryan Hooi
6. Trevor Darrell
7. Luc Van Gool
8. Mingkui Tan

Citations in the corpus

• Unleashing the Power of Contrastive Self-Supervised Visual Models via Contrast-Regularized Fine-Tuning (2021, background): “, 2019a), which has been found difficult to handle even in the most advanced AT methods (Zhang et al., 2021; Yang et al., 2020).”
• Diffusion Models for Adversarial Purification (2022, unknown): “Zhangetal.,2020) 59.64 59.96 (Wu et al., 2020) 60.04 62.11 (Gowal et al., 2020) 62.80 62.70 (Wu et al., 2020) 56.17 59.18 (Rebuffi et al., 2021) 60.75 61.72 (Gowal et al., 2021) 63.44 65.24 WideResNet-70-16 (Gowaletal.,2020) 65.88 66.02 (Rebuffi et al., 2021) 66.58 68.56 (Gowal et al., 2020) 57.20…”