Rethinking Transferable Adversarial Attacks on Point Clouds from a Compact Subspace Perspective
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866910006241132544 |
|---|---|
| author | Tang, Keke Liu, Xianheng Peng, Weilong Wang, Xiaofei Liu, Daizong Zhu, Peican Lu, Can Tian, Zhihong |
| author_facet | Tang, Keke Liu, Xianheng Peng, Weilong Wang, Xiaofei Liu, Daizong Zhu, Peican Lu, Can Tian, Zhihong |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_23102 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Rethinking Transferable Adversarial Attacks on Point Clouds from a Compact Subspace Perspective Tang, Keke Liu, Xianheng Peng, Weilong Wang, Xiaofei Liu, Daizong Zhu, Peican Lu, Can Tian, Zhihong Computer Vision and Pattern Recognition 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. |
| title | Rethinking Transferable Adversarial Attacks on Point Clouds from a Compact Subspace Perspective |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2601.23102 |