paper-with-me

홈 › Papers

What Do Adversarially trained Neural Networks Focus: A Fourier Domain-based Study

2022-03-16 · Binxiao Huang, Chaofan Tao, Rui Lin, Ngai Wong

Although many fields have witnessed the superior performance brought about by deep learning, the robustness of neural networks remains an open issue. Specifically, a small adversarial perturbation on the input may cause the model to produce a completely different output. Such poor robustness implies many potential hazards, especially in security-critical applications, e.g., autonomous driving and mobile robotics. This work studies what information the adversarially trained model focuses on. Empirically, we notice that the differences between the clean and adversarial data are mainly distributed in the low-frequency region. We then find that an adversarially-trained model is more robust than its naturally-trained counterpart due to the reason that the former pays more attention to learning the dominant information in low-frequency components. In addition, we consider two common ways to improve model robustness, namely, by data augmentation and by using stronger network architectures, and understand these techniques from a frequency-domain perspective. We are hopeful this work can shed light on the design of more robust neural networks.

📄 PDF Abstract BibTeX arXiv:2203.08739

Code (0)

등록된 구현이 없습니다.

Tasks

Autonomous DrivingData Augmentation

Similar Papers 제목 키워드 기반

Adversarially Trained Neural Policies in the Fourier Domain

2021-06-18 · ICML Workshop AML 2021 7 · Ezgi Korkmaz

Reinforcement learning policies based on deep neural networks are vulnerable to imperceptible adversarial perturbations to their inputs, in much the same way as neural network image classifiers. Recent work has proposed …

Deep Reinforcement Learningreinforcement-learningReinforcement LearningReinforcement Learning (RL)

Investigating Vulnerabilities of Deep Neural Policies

2021-08-30 · Ezgi Korkmaz

Reinforcement learning policies based on deep neural networks are vulnerable to imperceptible adversarial perturbations to their inputs, in much the same way as neural network image classifiers. Recent work has proposed …

Deep Reinforcement Learningreinforcement-learningReinforcement Learning (RL)Sensitivity

DeepDGA: Adversarially-Tuned Domain Generation and Detection

2016-10-06 · Hyrum S. Anderson, Jonathan Woodbridge, Bobby Filar

Many malware families utilize domain generation algorithms (DGAs) to establish command and control (C&C) connections. While there are many methods to pseudorandomly generate domains, we focus in this paper on detecting (…

BIG-bench Machine LearningDecoderDeep LearningGenerative Adversarial Network

Domain Generalisation via Domain Adaptation: An Adversarial Fourier Amplitude Approach

2023-02-23 · Minyoung Kim, Da Li, Timothy Hospedales

We tackle the domain generalisation (DG) problem by posing it as a domain adaptation (DA) task where we adversarially synthesise the worst-case target domain and adapt a model to that worst-case domain, thereby improving…

Domain Adaptation

What Do Adversarially Robust Models Look At?

2019-05-19 · Takahiro Itazuri, Yoshihiro Fukuhara, Hirokatsu Kataoka, Shigeo Morishima

In this paper, we address the open question: "What do adversarially robust models look at?" Recently, it has been reported in many works that there exists the trade-off between standard accuracy and adversarial robustnes…

Adversarial RobustnessOpen-Ended Question Answering