Transferable Unlearnable Examples

TL;DR

This paper introduces a new unlearnable example method based on Classwise Separability Discriminant (CSD) that enhances transferability across different training settings and datasets, addressing limitations of previous approaches.

Contribution

The paper proposes a novel unlearnable strategy using CSD to improve transferability of data protection across various training scenarios and datasets.

Findings

The CSD-based method significantly improves transferability of unlearnable effects.

Experiments show effective protection across multiple datasets and training settings.

The approach outperforms existing unlearnable techniques in transferability.

Abstract

With more people publishing their personal data online, unauthorized data usage has become a serious concern. The unlearnable strategies have been introduced to prevent third parties from training on the data without permission. They add perturbations to the users' data before publishing, which aims to make the models trained on the perturbed published dataset invalidated. These perturbations have been generated for a specific training setting and a target dataset. However, their unlearnable effects significantly decrease when used in other training settings and datasets. To tackle this issue, we propose a novel unlearnable strategy based on Classwise Separability Discriminant (CSD), which aims to better transfer the unlearnable effects to other training settings and datasets by enhancing the linear separability. Extensive experiments demonstrate the transferability of the proposed unlearnable examples across training settings and datasets.