Robust Alignment: Harmonizing Clean Accuracy and Adversarial Robustness in Adversarial Training

• URL: http://arxiv.org/abs/2604.26496v1
• Date: Wed, 29 Apr 2026 09:59:41 GMT
• Title: Robust Alignment: Harmonizing Clean Accuracy and Adversarial Robustness in Adversarial Training
• Abstract summary: Adversarial Training (AT) is one of the most effective methods for developing robust deep neural networks (DNNs)<n>We show that Varying input perturbation intensities for training samples near decision boundaries in AT have minimal impact on model robustness.<n>We propose a new Robust Alignment Adversarial Training (RAAT) method, effectively harmonizing accuracy and robustness.
• Score: 16.873607054158793
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Adversarial Training (AT) is one of the most effective methods for developing robust deep neural networks (DNNs). However, AT faces a trade-off problem between clean accuracy and adversarial robustness. In this work, we reveal a surprising phenomenon for the first time: Varying input perturbation intensities for training samples near decision boundaries in AT have minimal impact on model robustness. This finding directly exposes the inconsistency between accuracy and robustness score fluctuations, leading us to identify the misalignment between input and latent spaces as a critical driver of the robustness-accuracy trade-off. To mitigate this misalignment for harmonizing accuracy and robustness, we define Robust Alignment as a new AT target, encouraging the model perception to change with input perturbations provided the final label prediction remains unchanged, which can be achieved via two novel ideas. First, we suggest a reduced and fixed perturbation intensity for those boundary samples, which facilitates the model to utilize the perturbations as learnable patterns, instead of noises that complicate decision boundaries meaninglessly. Second, we propose a Domain Interpolation Consistency Adversarial Regularization (DICAR), based on rigorous theoretical derivations, which explicitly introduces semantic alignment between input and latent spaces into AT. Based on these two ideas, we end up with a new Robust Alignment Adversarial Training (RAAT) method, effectively harmonizing accuracy and robustness. Extensive experiments on CIFAR-10, CIFAR-100, and Tiny-ImageNet with ResNet-18, PreActResNet-18, and WideResNet-28-10 demonstrate the effectiveness of RAAT in improving the trade-off beyond four common baselines and a total of 14 related state-of-the-art (SOTA) works.

Related papers

• Mitigating Error Amplification in Fast Adversarial Training [58.74042726356826]
  Fast Adversarial Training (FAT) has proven effective in enhancing model robustness by encouraging networks to learn perturbation-invariant representations.<n>FAT often suffers from catastrophic overfitting (CO), where the model overfits to the training attack and fails to generalize to unseen ones.<n>We propose a Distribution-aware Dynamic Guidance (DDG) strategy that dynamically adjusts both the perturbation budget and supervision signal.
  arXiv  Detail & Related papers  (2026-04-27T11:23:18Z)
• The Geometry of Robustness: Optimizing Loss Landscape Curvature and Feature Manifold Alignment for Robust Finetuning of Vision-Language Models [33.10473765967503]
  Generalization-preserving methods retain ID/OOD performance but leave models vulnerable to adversarial attacks.<n>Our key insight is that the robustness trade-off stems from two geometric failures: sharp, anisotropic minima in parameter space and unstable feature representations that deform under perturbation.<n>We propose GRACE, a unified fine-tuning framework that jointly regularizes the parameter-space curvature and feature-space invariance for VLMs.
  arXiv  Detail & Related papers  (2026-03-28T05:22:00Z)
• Adaptive Dual Uncertainty Optimization: Boosting Monocular 3D Object Detection under Test-Time Shifts [80.32933059529135]
  Test-Time Adaptation (TTA) methods have emerged to adapt to target distributions during inference.<n>We propose Dual Uncertainty Optimization (DUO), the first TTA framework designed to jointly minimize both uncertainties for robust M3OD.<n>In parallel, we design a semantic-aware normal field constraint that preserves geometric coherence in regions with clear semantic cues.
  arXiv  Detail & Related papers  (2025-08-28T07:09:21Z)
• Failure Cases Are Better Learned But Boundary Says Sorry: Facilitating Smooth Perception Change for Accuracy-Robustness Trade-Off in Adversarial Training [6.312209034178382]
  Adversarial Training (AT) is one of the most effective methods to train robust Deep Neural Networks (DNNs)<n>AT creates an inherent trade-off between clean accuracy and adversarial robustness.<n>We propose a novel Robust Perception Adversarial Training (RPAT) method, effectively mitigating the current accuracy-robustness trade-off.
  arXiv  Detail & Related papers  (2025-08-04T08:31:45Z)
• New Paradigm of Adversarial Training: Releasing Accuracy-Robustness Trade-Off via Dummy Class [10.937432407005124]
  Adversarial Training (AT) is one of the most effective methods to enhance the robustness of Deep Neural Networks (DNNs)<n>Existing AT methods suffer from an inherent accuracy-robustness trade-off.<n>We propose a new AT paradigm by introducing an additional dummy class for each original class.
  arXiv  Detail & Related papers  (2024-10-16T15:36:10Z)
• Towards Continual Learning Desiderata via HSIC-Bottleneck Orthogonalization and Equiangular Embedding [55.107555305760954]
  We propose a conceptually simple yet effective method that attributes forgetting to layer-wise parameter overwriting and the resulting decision boundary distortion. Our method achieves competitive accuracy performance, even with absolute superiority of zero exemplar buffer and 1.02x the base model.
  arXiv  Detail & Related papers  (2024-01-17T09:01:29Z)
• Towards Calibrated Robust Fine-Tuning of Vision-Language Models [97.19901765814431]
  This work proposes a robust fine-tuning method that improves both OOD accuracy and confidence calibration simultaneously in vision language models. We show that both OOD classification and OOD calibration errors have a shared upper bound consisting of two terms of ID data. Based on this insight, we design a novel framework that conducts fine-tuning with a constrained multimodal contrastive loss enforcing a larger smallest singular value.
  arXiv  Detail & Related papers  (2023-11-03T05:41:25Z)
• Frequency Regularization for Improving Adversarial Robustness [8.912245110734334]
  adversarial training (AT) has proven to be an effective defense approach. We propose a frequency regularization (FR) to align the output difference in the spectral domain. We find that our method achieves the strongest robustness against attacks by PGD-20, C&W and Autoattack.
  arXiv  Detail & Related papers  (2022-12-24T13:14:45Z)
• Robustness and Accuracy Could Be Reconcilable by (Proper) Definition [109.62614226793833]
  The trade-off between robustness and accuracy has been widely studied in the adversarial literature. We find that it may stem from the improperly defined robust error, which imposes an inductive bias of local invariance. By definition, SCORE facilitates the reconciliation between robustness and accuracy, while still handling the worst-case uncertainty.
  arXiv  Detail & Related papers  (2022-02-21T10:36:09Z)
• SmoothMix: Training Confidence-calibrated Smoothed Classifiers for Certified Robustness [61.212486108346695]
  We propose a training scheme, coined SmoothMix, to control the robustness of smoothed classifiers via self-mixup. The proposed procedure effectively identifies over-confident, near off-class samples as a cause of limited robustness. Our experimental results demonstrate that the proposed method can significantly improve the certified $ell$-robustness of smoothed classifiers.
  arXiv  Detail & Related papers  (2021-11-17T18:20:59Z)
• Uncertainty-Aware Deep Calibrated Salient Object Detection [74.58153220370527]
  Existing deep neural network based salient object detection (SOD) methods mainly focus on pursuing high network accuracy. These methods overlook the gap between network accuracy and prediction confidence, known as the confidence uncalibration problem. We introduce an uncertaintyaware deep SOD network, and propose two strategies to prevent deep SOD networks from being overconfident.
  arXiv  Detail & Related papers  (2020-12-10T23:28:36Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.