Existing adversarial attacks on the point cloud data involve generating fake obstacles, removing objects or producing fake predictions. Despite the demonstrated success, these approaches have three limitations. First, manipulating point data, which was originally designed for point-based representation, is ...
Decoupling Direction and Norm for Efficient Gradient-Based Adversarial Attacks and Defenses 说在前面 1.提出的问题 2.提出的方法 2.1 相关工作 2.2 算法介绍 3.实验结果 3.1 Untargeted Attack 3.2 Targeted Attack 3.3 Defense Evaluation 4.结论 Decoupling Direction and Norm for Efficient Gradient-Based L2 Ad...
Among these attacks, methods based on gradient information are widely used due to their simplicity and efficiency. However, existing attack methods are often prone to local optimality. In this paper, we propose a label poisoning method that incorporates gradient and node importance scores. The ...
In this paper,we f ind that existing IG-based attacks have limited trans-ferability due to their naive adoption of IG in model inter-pretability. To address this limitation, we focus on the IG inte-gration path and ref ine it in three aspects: multiplicity, mono-tonicity, and diversity,...
behavior of human–machine interactions. It provides a promising cyber security layer against password leaking and network attacks for the protection of personal information. In addition, the AIOM touch sensor can be functionalized as a linear interactive control interface or a circular touch panel for...
The cloud makes data storage more accessible and adaptable, but unwanted attacks and operations still exist. Sensitive data could be covertly stored on the cloud server. Data security is essential as a result. Here, we combine Stochastic Gradient Descent long short-term memory (SGD-LSTM) with ...
Particularly, the number of vulnerable or unprotected IoT devices has drastically increased, along with the amount of suspicious activity, such as IoT botnet and large-scale cyber-attacks. In order to address this security issue, researchers have deployed machine and deep learning methods to detect ...
et al. Security constrained unit commitment in smart energy systems: A flexibility-driven approach considering false data injection attacks in electric vehicle parking lots. International Journal of Electrical Power and Energy Systems, 2024, 161: 110180. DOI:10.1016/j.ijepes.2024.110180 74. Rahman,...
(Pathak, Lu, Hunt, Girvan, & Ott, 2017), for implementing reservoir based observers for spatio temporal complex systems (Lu et al., 2017, Zimmermann and Parlitz, 2018), for detection of smart grid attacks (Hamedani, Liu, Atat, Wu, & Yi, 2018) and for digit recognition (Jalalvand, ...
28 Oct 2021·Lifan Yuan,Yichi Zhang,Yangyi Chen,Wei Wei· Despite recent success on various tasks, deep learning techniques still perform poorly on adversarial examples with small perturbations. While optimization-based methods for adversarial attacks are well-explored in the field of computer vision...