Identifying and Understanding Cross-Class Features in Adversarial Training

• URL: http://arxiv.org/abs/2506.05032v1
• Date: Thu, 05 Jun 2025 13:40:11 GMT
• Title: Identifying and Understanding Cross-Class Features in Adversarial Training
• Authors: Zeming Wei, Yiwen Guo, Yisen Wang,
• Abstract summary: Adversarial training (AT) has been considered one of the most effective methods for making deep neural networks robust against adversarial attacks.<n>We present a novel perspective on studying AT through the lens of class-wise feature attribution.
• Score: 40.939746601347785
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Adversarial training (AT) has been considered one of the most effective methods for making deep neural networks robust against adversarial attacks, while the training mechanisms and dynamics of AT remain open research problems. In this paper, we present a novel perspective on studying AT through the lens of class-wise feature attribution. Specifically, we identify the impact of a key family of features on AT that are shared by multiple classes, which we call cross-class features. These features are typically useful for robust classification, which we offer theoretical evidence to illustrate through a synthetic data model. Through systematic studies across multiple model architectures and settings, we find that during the initial stage of AT, the model tends to learn more cross-class features until the best robustness checkpoint. As AT further squeezes the training robust loss and causes robust overfitting, the model tends to make decisions based on more class-specific features. Based on these discoveries, we further provide a unified view of two existing properties of AT, including the advantage of soft-label training and robust overfitting. Overall, these insights refine the current understanding of AT mechanisms and provide new perspectives on studying them. Our code is available at https://github.com/PKU-ML/Cross-Class-Features-AT.

Related papers

• Shortcut Learning Susceptibility in Vision Classifiers [3.004632712148892]
  Shortcut learning is where machine learning models exploit spurious correlations in data instead of capturing meaningful features.<n>This phenomenon is prevalent across various machine learning applications, including vision, natural language processing, and speech recognition.<n>We systematically evaluate these architectures by introducing deliberate shortcuts into the dataset that are positionally correlated with class labels.
  arXiv  Detail & Related papers  (2025-02-13T10:25:52Z)
• Intra-task Mutual Attention based Vision Transformer for Few-Shot Learning [12.5354658533836]
  Humans possess remarkable ability to accurately classify new, unseen images after being exposed to only a few examples. For artificial neural network models, determining the most relevant features for distinguishing between two images with limited samples presents a challenge. We propose an intra-task mutual attention method for few-shot learning, that involves splitting the support and query samples into patches.
  arXiv  Detail & Related papers  (2024-05-06T02:02:57Z)
• Multi-Modal Prompt Learning on Blind Image Quality Assessment [65.0676908930946]
  Image Quality Assessment (IQA) models benefit significantly from semantic information, which allows them to treat different types of objects distinctly. Traditional methods, hindered by a lack of sufficiently annotated data, have employed the CLIP image-text pretraining model as their backbone to gain semantic awareness. Recent approaches have attempted to address this mismatch using prompt technology, but these solutions have shortcomings. This paper introduces an innovative multi-modal prompt-based methodology for IQA.
  arXiv  Detail & Related papers  (2024-04-23T11:45:32Z)
• Continual Learning with Pre-Trained Models: A Survey [61.97613090666247]
  Continual Learning aims to overcome the catastrophic forgetting of former knowledge when learning new ones. This paper presents a comprehensive survey of the latest advancements in PTM-based CL.
  arXiv  Detail & Related papers  (2024-01-29T18:27:52Z)
• Enhancing Robust Representation in Adversarial Training: Alignment and Exclusion Criteria [61.048842737581865]
  We show that Adversarial Training (AT) omits to learning robust features, resulting in poor performance of adversarial robustness. We propose a generic framework of AT to gain robust representation, by the asymmetric negative contrast and reverse attention. Empirical evaluations on three benchmark datasets show our methods greatly advance the robustness of AT and achieve state-of-the-art performance.
  arXiv  Detail & Related papers  (2023-10-05T07:29:29Z)
• Salient Feature Extractor for Adversarial Defense on Deep Neural Networks [2.993911699314388]
  Motivated by the observation that adversarial examples are due to the non-robust feature learned from the original dataset by models, we propose the concepts of salient feature(SF) and trivial feature(TF) We put forward a novel detection and defense method named salient feature extractor (SFE) to defend against adversarial attacks.
  arXiv  Detail & Related papers  (2021-05-14T12:56:06Z)
• ConCAD: Contrastive Learning-based Cross Attention for Sleep Apnea Detection [16.938983046369263]
  We propose a contrastive learning-based cross attention framework for sleep apnea detection (named ConCAD) Our proposed framework can be easily integrated into standard deep learning models to utilize expert knowledge and contrastive learning to boost performance.
  arXiv  Detail & Related papers  (2021-05-07T02:38:56Z)
• Fine-grained Angular Contrastive Learning with Coarse Labels [72.80126601230447]
  We introduce a novel 'Angular normalization' module that allows to effectively combine supervised and self-supervised contrastive pre-training. This work will help to pave the way for future research on this new, challenging, and very practical topic of C2FS classification.
  arXiv  Detail & Related papers  (2020-12-07T08:09:02Z)
• Symbiotic Adversarial Learning for Attribute-based Person Search [86.7506832053208]
  We present a symbiotic adversarial learning framework, called SAL.Two GANs sit at the base of the framework in a symbiotic learning scheme. Specifically, two different types of generative adversarial networks learn collaboratively throughout the training process.
  arXiv  Detail & Related papers  (2020-07-19T07:24:45Z)

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.