Adversarial attacks significantly challenge the safe deployment of deep learning models, particularly in real-world applications. Traditional defenses often rely on computationally intensive optimization (e.g., adversarial training or data augmentation) to improve robustness, whereas the human visual system achieves inherent robustness to adversarial perturbations through evolved biological mechanisms. We hypothesize that attention guided non-homogeneous sparse sampling and predictive coding plays a key role in this robustness. To test this hypothesis, we propose a novel defense framework incorporating three key biological mechanisms: foveal-peripheral processing, saccadic eye movements, and cortical filling-in. Our approach employs reinforcement learning-guided saccades to selectively capture multiple foveal-peripheral glimpses, which are integrated into a reconstructed image before classification. This biologically inspired preprocessing effectively mitigates adversarial noise, preserves semantic integrity, and notably requires no retraining or fine-tuning of downstream classifiers, enabling seamless integration with existing systems. Experiments on the ImageNet dataset demonstrate that our method improves system robustness across diverse classifiers and attack types, while significantly reducing training overhead compared to both biologically and non-biologically inspired defense techniques.

This paper studies the adversarial robustness of conformal novelty detection. In particular, we focus on two powerful learning-based frameworks that come with finite-sample false discovery rate (FDR) control: one is AdaDetect (by Marandon et al., 2024) that is based on the positive-unlabeled classifier, and the other is a one-class classifier-based approach (by Bates et al., 2023). While they provide rigorous statistical guarantees under benign conditions, their behavior under adversarial perturbations remains underexplored. We first formulate an oracle attack setup, under the AdaDetect formulation, that quantifies the worst-case degradation of FDR, deriving an upper bound that characterizes the statistical cost of attacks. This idealized formulation directly motivates a practical and effective attack scheme that only requires query access to the output labels of both frameworks. Coupling these formulations with two popular and complementary black-box adversarial algorithms, we systematically evaluate the vulnerability of both frameworks on synthetic and real-world datasets. Our results show that adversarial perturbations can significantly increase the FDR while maintaining high detection power, exposing fundamental limitations of current error-controlled novelty detection methods and motivating the development of more robust alternatives.

Federated learning (FL) enables privacy-preserving model training by keeping data decentralized. However, it remains vulnerable to label-flipping attacks, where malicious clients manipulate labels to poison the global model. Despite their simplicity, these attacks can severely degrade model performance, and defending against them remains challenging. We introduce AntiFLipper, a novel and computationally efficient defense against multi-class label-flipping attacks in FL. Unlike existing methods that ensure security at the cost of high computational overhead, AntiFLipper employs a novel client-side detection strategy, significantly reducing the central server's burden during aggregation. Comprehensive empirical evaluations across multiple datasets under different distributions demonstrate that AntiFLipper achieves accuracy comparable to state-of-the-art defenses while requiring substantially fewer computational resources in server side. By balancing security and efficiency, AntiFLipper addresses a critical gap in existing defenses, making it particularly suitable for resource-constrained FL deployments where both model integrity and operational efficiency are essential.