14 May 2024 in IEEE Transactions on Intelligent Transportation Systems
Traffic accidents caused by driver distractions have seriously endangered public safety, with driver distractions typically stemming from behaviors beyond safe driving. Recently, vision-based driver behavior recognition has attracted much attention, achieving great success with deep learning-based schemes. However, the generalization ability of these models in real-world scenarios remains unsatisfactory. In this paper, we conduct an in-depth investigation into the underlying causes of this unsatisfactory generalization and conclude that the behavior features extracted by convolutional neural networks are intertwined with driver identity features. Based on this discovery, we propose a feature decomposition (FD) framework to disentangle these two types of features. The separated behavior features, referred to as driver-irrelevant behavior features, are subsequently leveraged for behavior recognition. Moreover, we introduce a co-training strategy to optimize the FD framework. This strategy enables behavior features and identity features to provide mutual auxiliary signals and encourages each other to drop the information that do not belong to them, so that the learned behavior features can be driver-irrelevant. Rigorous experiments are conducted on two widely-studied datasets, yielding results that demonstrate the superior performance of our framework and its improved generalization capabilities. Importantly, our framework’s fast inference capabilities make it highly suitable for real-world scenarios. Codes are released at https://github.com/gaohangcodes/ LearningDriverIrrelevantFeatures4DBR.
17 July 2025 in Journal of King Saud University Computer and Information Sciences
Federated learning has emerged as a promising paradigm for privacy-preserving multi-source data fusion. However, its distributed nature makes it vulnerable to poisoning attacks. Malicious clients inject poisoned noises into their local models, severely degrading the global model’s performance. While existing defense mechanisms attempt to mitigate these threats, they often overlook a critical vulnerability: historical knowledge can reveal clients’ training trajectories or data distributions, providing adversaries with new opportunities for attack. At the same time, the historical knowledge also serves as a tool for detecting and neutralizing malicious activities. In this paper, we address the challenge by proposing a novel attack method and a robust defense framework. First, we propose HisMSA, a stealthy, untargeted model poisoning attack based on historical information perception. HisMSA requires only a few compromised clients to significantly degrade global model accuracy while evading existing defenses. Second, we present DynHisFL, a versatile defense framework that enhances the global model’s robustness by identifying and filtering poisoned models through the analysis of divergence in historical model distributions, making it adaptable to various untargeted attacks. The experiment shows that with 10% malicious clients, the accuracy of HisMSA drops to 30%, approaching unavailability. Despite defenses like statistical feature analysis and topology distance detection, HisMSA still degrades performance by 5%-20%. DynHisFL improves accuracy by 71.3% even with a 20% attack ratio, while achieving an outstanding 98.7% poisoning detection accuracy. Ten model distribution metrics are evaluated, with Maximum Mean Discrepancy achieving a recall rate of 70%-100% across various poisoning strategies.
11 December 2025 in Applied Soft Computing
20 May 2025 in Journal of King Saud University Computer and Information Sciences
While large data owners outsource their data to cloud for storage, this results in the security issues of data of all users. To provide users the means to verify whether the data is correctly and accurately stored in the cloud server or not, people have suggested many feasible scheme for data auditing. However, to cope with the problem of cloud server is still an urgently needs to be solved, therefore, we proposed a novel data auditing scheme of cloud storage for secret shares generated by sharing matrix, which is seldom discussed in present literatures. In the proposed scheme, the data owner first generates some secret shares through \(\left( k,n\right) \) sharing matrix, and then outsources these secret shares to different or the same cloud servers, then the skillfully designed protocols are provided to verify the data integrity of the cloud servers of different position, and the proposed protocols also support data dynamics updating. Moreover, mathematical analysis shows that the integrity verification of secret shares in cloud server can be obtained by the collaboration fulfillment of data verification in other cloud servers. Security analyses are given to assure that the proposed scheme can guarantee the privacy and security of data. The discussions and comparisons with some available methods are presented to show that lower computation cost and better performance are obtained for the proposed protocols.