Rethinking Transferable Adversarial Attacks on Point Clouds from a Compact Subspace Perspective

2026-01-30

AI Summary

Existing transferable adversarial attack methods on point clouds are limited by model-specific gradients or heuristic strategies, resulting in poor generalization. This work proposes CoSA, a novel framework that views point clouds as low-rank combinations of class prototypes within a compact semantic subspace. By optimizing perturbations in this shared subspace, CoSA suppresses model-dependent noise and steers the adversarial examples toward semantically consistent directions. The approach enhances cross-architecture transferability through class-specific prototype modeling, low-rank subspace constraints, and preservation of semantic structure. Extensive experiments demonstrate that CoSA outperforms state-of-the-art methods across multiple datasets and network architectures, while maintaining high imperceptibility and robustness against mainstream defense strategies.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness
• Multiagent Systems: Adversarial Agents

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Security and Privacy: Data transparency and provenance
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Transferable adversarial attacks on point clouds remain challenging, as existing methods often rely on model-specific gradients or heuristics that limit generalization to unseen architectures. In this paper, we rethink adversarial transferability from a compact subspace perspective and propose CoSA, a transferable attack framework that operates within a shared low-dimensional semantic space. Specifically, each point cloud is represented as a compact combination of class-specific prototypes that capture shared semantic structure, while adversarial perturbations are optimized within a low-rank subspace to induce coherent and architecture-agnostic variations. This design suppresses model-dependent noise and constrains perturbations to semantically meaningful directions, thereby improving cross-model transferability without relying on surrogate-specific artifacts. Extensive experiments on multiple datasets and network architectures demonstrate that CoSA consistently outperforms state-of-the-art transferable attacks, while maintaining competitive imperceptibility and robustness under common defense strategies. Codes will be made public upon paper acceptance.

Problem

• transferable adversarial attacks
• point clouds
• adversarial transferability
• model generalization
• cross-model attack

Innovation

• compact subspace
• transferable adversarial attack
• point clouds
• low-rank perturbation
• semantic prototypes