SPBA is a curvature-aware patch-wise backdoor attack framework for 3D point clouds that achieves high stealthiness and reduces trigger computation by 98.43% compared to sample-wise baselines while maintaining attack performance.
Point cloud matters: Rethinking the impact of different observation spaces on robot learning.Ad- vances in Neural Information Processing Systems (NeurIPS), 37:77799–77830
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Stealthy Patch-Wise Backdoor Attack in 3D Point Cloud via Curvature Awareness
SPBA is a curvature-aware patch-wise backdoor attack framework for 3D point clouds that achieves high stealthiness and reduces trigger computation by 98.43% compared to sample-wise baselines while maintaining attack performance.