Non-Robust Features are Not Always Useful in One-Class Classification

• URL: http://arxiv.org/abs/2407.06372v1
• Date: Mon, 8 Jul 2024 20:32:19 GMT
• Title: Non-Robust Features are Not Always Useful in One-Class Classification
• Abstract summary: The robustness of machine learning models has been questioned by the existence of adversarial examples. We examine the threat of adversarial examples in practical applications that require lightweight models for one-class classification.
• Score: 19.533796635619613
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The robustness of machine learning models has been questioned by the existence of adversarial examples. We examine the threat of adversarial examples in practical applications that require lightweight models for one-class classification. Building on Ilyas et al. (2019), we investigate the vulnerability of lightweight one-class classifiers to adversarial attacks and possible reasons for it. Our results show that lightweight one-class classifiers learn features that are not robust (e.g. texture) under stronger attacks. However, unlike in multi-class classification (Ilyas et al., 2019), these non-robust features are not always useful for the one-class task, suggesting that learning these unpredictive and non-robust features is an unwanted consequence of training.

Related papers

• CAARMA: Class Augmentation with Adversarial Mixup Regularization [34.02819618734268]
  CAARMA is a class augmentation framework for speaker verification. We introduce synthetic classes through data mixing in the embedding space, expanding the number of training classes. We evaluate CAARMA on multiple speaker verification tasks, as well as other representative zero-shot comparison-based speech analysis tasks.
  arXiv  Detail & Related papers  (2025-03-20T21:41:16Z)
• Towards Class-wise Robustness Analysis [15.351461000403074]
  Exploiting weakly robust classes is a potential avenue for attackers to fool the image recognition models. This study investigates class-to-class biases across adversarially trained robust classification models. We find that the number of false positives of classes as specific target classes significantly impacts their vulnerability to attacks.
  arXiv  Detail & Related papers  (2024-11-29T17:09:59Z)
• Few-Shot Class-Incremental Learning via Training-Free Prototype Calibration [67.69532794049445]
  We find a tendency for existing methods to misclassify the samples of new classes into base classes, which leads to the poor performance of new classes. We propose a simple yet effective Training-frEE calibratioN (TEEN) strategy to enhance the discriminability of new classes.
  arXiv  Detail & Related papers  (2023-12-08T18:24:08Z)
• Adversarial Examples Are Not Real Features [55.5802648842869]
  A well-known theory by citetilyas 2019adversarial explains adversarial vulnerability from a data perspective. We re-examine the theory from a larger context by incorporating multiple learning paradigms.
  arXiv  Detail & Related papers  (2023-10-29T08:50:27Z)
• 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)
• Which Features are Learnt by Contrastive Learning? On the Role of Simplicity Bias in Class Collapse and Feature Suppression [59.97965005675144]
  Contrastive learning (CL) has emerged as a powerful technique for representation learning, with or without label supervision. We provide the first unified theoretically rigorous framework to determine textitwhich features are learnt by CL. We present increasing embedding dimensionality and improving the quality of data augmentations as two theoretically motivated solutions.
  arXiv  Detail & Related papers  (2023-05-25T23:37:22Z)
• Generalization Bounds for Few-Shot Transfer Learning with Pretrained Classifiers [26.844410679685424]
  We study the ability of foundation models to learn representations for classification that are transferable to new, unseen classes. We show that the few-shot error of the learned feature map on new classes is small in case of class-feature-variability collapse.
  arXiv  Detail & Related papers  (2022-12-23T18:46:05Z)
• Towards A Conceptually Simple Defensive Approach for Few-shot classifiers Against Adversarial Support Samples [107.38834819682315]
  We study a conceptually simple approach to defend few-shot classifiers against adversarial attacks. We propose a simple attack-agnostic detection method, using the concept of self-similarity and filtering. Our evaluation on the miniImagenet (MI) and CUB datasets exhibit good attack detection performance.
  arXiv  Detail & Related papers  (2021-10-24T05:46:03Z)
• Prototypical Classifier for Robust Class-Imbalanced Learning [64.96088324684683]
  We propose textitPrototypical, which does not require fitting additional parameters given the embedding network. Prototypical produces balanced and comparable predictions for all classes even though the training set is class-imbalanced. We test our method on CIFAR-10LT, CIFAR-100LT and Webvision datasets, observing that Prototypical obtains substaintial improvements compared with state of the arts.
  arXiv  Detail & Related papers  (2021-10-22T01:55:01Z)
• Localized Uncertainty Attacks [9.36341602283533]
  We present localized uncertainty attacks against deep learning models. We create adversarial examples by perturbing only regions in the inputs where a classifier is uncertain. Unlike $ell_p$ ball or functional attacks which perturb inputs indiscriminately, our targeted changes can be less perceptible.
  arXiv  Detail & Related papers  (2021-06-17T03:07:22Z)
• Learning and Evaluating Representations for Deep One-class Classification [59.095144932794646]
  We present a two-stage framework for deep one-class classification. We first learn self-supervised representations from one-class data, and then build one-class classifiers on learned representations. In experiments, we demonstrate state-of-the-art performance on visual domain one-class classification benchmarks.
  arXiv  Detail & Related papers  (2020-11-04T23:33:41Z)
• Adversarial Examples and Metrics [14.068394742881425]
  Adversarial examples are a type of attack on machine learning (ML) systems which cause misclassification of inputs. We study the limitations of robust classification if the target metric is uncertain.
  arXiv  Detail & Related papers  (2020-07-14T12:20:53Z)
• Leveraging Siamese Networks for One-Shot Intrusion Detection Model [0.0]
  Supervised Machine Learning (ML) to enhance Intrusion Detection Systems has been the subject of significant research. retraining the models in-situ renders the network susceptible to attacks owing to the time-window required to acquire a sufficient volume of data. Here, a complementary approach referred to as 'One-Shot Learning', whereby a limited number of examples of a new attack-class is used to identify a new attack-class. A Siamese Network is trained to differentiate between classes based on pairs similarities, rather than features, allowing to identify new and previously unseen attacks.
  arXiv  Detail & Related papers  (2020-06-27T11:40:01Z)

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.