analiza aiqbench
This research introduces Elastic Dictionary Learning Networks (EDLNets), a ResNet architecture designed to close the gap between noise robustness and adversarial defense in deep learning models. By integrating dictionary structure priors, the study demonstrates significant improvements in generalization and resilience against strong adaptive attacks across various datasets and threat models. This work establishes a new direction for enhancing model security without sacrificing performance, addressing a critical vulnerability in current CNN designs.
Boosting Adversarial Robustness and Generalization with Dictionary Structure
arXiv:2502.00834v2 Announce Type: replace Abstract: This work investigates a novel approach to boost adversarial robustness and generalization by incorporating structural prior into the design of deep learning models. Specifically, our study surprisingly reveals that existing dictionary learning-inspired convolutional neural networks (CNNs) are robust against random noise but remain highly vulnerable to adversarial attacks. To address this, we propose Elastic Dictionary Learning Networks (EDLNets), a novel ResNet architecture that significantly enhances adversarial robustness and generalization. Extensive and reliable experiments demonstrate consistent improvements in adversarial robustness across multiple datasets, backbone architectures, and threat models. To the best of our knowledge, this is the first work to discover and validate that dictionary structure can reliably enhance deep learning robustness under strong adaptive attacks, unveiling a promising direction for future research.
Abstract:This work investigates a novel approach to boost adversarial robustness and generalization by incorporating structural prior into the design of deep learning models. Specifically, our study surprisingly reveals that existing dictionary learning-inspired convolutional neural networks (CNNs) are robust against random noise but remain highly vulnerable to adversarial attacks. To address this, we propose Elastic Dictionary Learning Networks (EDLNets), a novel ResNet architecture that significantly enhances adversarial robustness and generalization. Extensive and reliable experiments demonstrate consistent improvements in adversarial robustness across multiple datasets, backbone architectures, and threat models. To the best of our knowledge, this is the first work to discover and validate that dictionary structure can reliably enhance deep learning robustness under strong adaptive attacks, unveiling a promising direction for future research.
Machine Learning (cs.LG); Cryptography and Security (cs.CR); Neural and Evolutionary Computing (cs.NE)
arXiv:2502.00834 [cs.LG]
(or arXiv:2502.00834v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2502.00834 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 223 cuvinte
Subiecte conexe
Problems of Proliferation
LessWrong · 29 sept. 2026
Angles of attack for continual learning safety
LessWrong · 16 iun. 2026
Synthesis and Deployment of Maximal Robust Control Barrier Functions through Adversarial Reinforcement Learning
arXiv · 21 sept. 2026
Neural Variability Enhances Artificial Network Robustness
arXiv · 15 iun. 2026