Unlearnable 3D Point Clouds: Class-wise Transformation Is All You Need

TL;DR

This work proposes the first integral unlearnable framework for 3D point clouds including two processes: an unlearnable data protection scheme, involving a class-wise setting established by a category-adaptive allocation strategy and multi-transformations assigned to samples, and a data restoration scheme that utilizes class-wise inverse matrix transformation.

Abstract

Traditional unlearnable strategies have been proposed to prevent unauthorized users from training on the 2D image data. With more 3D point cloud data containing sensitivity information, unauthorized usage of this new type data has also become a serious concern. To address this, we propose the first integral unlearnable framework for 3D point clouds including two processes: (i) we propose an unlearnable data protection scheme, involving a class-wise setting established by a category-adaptive allocation strategy and multi-transformations assigned to samples; (ii) we propose a data restoration scheme that utilizes class-wise inverse matrix transformation, thus enabling authorized-only training for unlearnable data. This restoration process is a practical issue overlooked in most existing unlearnable literature, \ie, even authorized users struggle to gain knowledge from 3D unlearnable data. Both theoretical and empirical results (including 6 datasets, 16 models, and 2 tasks) demonstrate the effectiveness of our proposed unlearnable framework. Our code is available at \url{https://github.com/CGCL-codes/UnlearnablePC}

Figures (10)

• Figure 1: An overview of existing seven types of 3D transformations. "*" denotes rigid transformations that do not alter the shape of the point cloud samples, while the remaining transformations are non-rigid transformations.
• Figure 2: (a) Training on the transformed ModelNet10 dataset (employing sample-wise, dataset-wise, and class-wise patterns) using PointNet classifier. (b) The high-level overview of the class-wise setting.
• Figure 3: An overview of our proposed integral unlearnable pipeline
• Figure 4: The test accuracy (%) results obtained after training on the clean, UMT, and restoration datasets.
• Figure 5: Hyper-parameter sensitivity analysis: The impact of hyperparameters $r_s$, $r_p$, $b_l$, and $b_u$ on the test accuracy results (%) on the UMT (using $\mathcal{R}\mathcal{S}$) ModelNet10 dataset.