Beyond the Patch: Exploring Vulnerabilities of Visuomotor Policies via Viewpoint-Consistent 3D Adversarial Object
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_ | 1866910042209386496 |
|---|---|
| author | Lee, Chanmi Yoon, Minsung Kim, Woojae Lee, Sebin Yoon, Sung-eui |
| author_facet | Lee, Chanmi Yoon, Minsung Kim, Woojae Lee, Sebin Yoon, Sung-eui |
| contents | Neural network-based visuomotor policies enable robots to perform manipulation tasks but remain susceptible to perceptual attacks. For example, conventional 2D adversarial patches are effective under fixed-camera setups, where appearance is relatively consistent; however, their efficacy often diminishes under dynamic viewpoints from moving cameras, such as wrist-mounted setups, due to perspective distortions. To proactively investigate potential vulnerabilities beyond 2D patches, this work proposes a viewpoint-consistent adversarial texture optimization method for 3D objects through differentiable rendering. As optimization strategies, we employ Expectation over Transformation (EOT) with a Coarse-to-Fine (C2F) curriculum, exploiting distance-dependent frequency characteristics to induce textures effective across varying camera-object distances. We further integrate saliency-guided perturbations to redirect policy attention and design a targeted loss that persistently drives robots toward adversarial objects. Our comprehensive experiments show that the proposed method is effective under various environmental conditions, while confirming its black-box transferability and real-world applicability. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_04913 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Beyond the Patch: Exploring Vulnerabilities of Visuomotor Policies via Viewpoint-Consistent 3D Adversarial Object Lee, Chanmi Yoon, Minsung Kim, Woojae Lee, Sebin Yoon, Sung-eui Robotics Computer Vision and Pattern Recognition Neural network-based visuomotor policies enable robots to perform manipulation tasks but remain susceptible to perceptual attacks. For example, conventional 2D adversarial patches are effective under fixed-camera setups, where appearance is relatively consistent; however, their efficacy often diminishes under dynamic viewpoints from moving cameras, such as wrist-mounted setups, due to perspective distortions. To proactively investigate potential vulnerabilities beyond 2D patches, this work proposes a viewpoint-consistent adversarial texture optimization method for 3D objects through differentiable rendering. As optimization strategies, we employ Expectation over Transformation (EOT) with a Coarse-to-Fine (C2F) curriculum, exploiting distance-dependent frequency characteristics to induce textures effective across varying camera-object distances. We further integrate saliency-guided perturbations to redirect policy attention and design a targeted loss that persistently drives robots toward adversarial objects. Our comprehensive experiments show that the proposed method is effective under various environmental conditions, while confirming its black-box transferability and real-world applicability. |
| title | Beyond the Patch: Exploring Vulnerabilities of Visuomotor Policies via Viewpoint-Consistent 3D Adversarial Object |
| topic | Robotics Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.04913 |