CanDO：基于无导数优化从姿态偏移中恢复黑盒点云模型
Recovering Black-Box Point Cloud Models from Pose Shifts via Derivative-Free Optimization
AI 导读
研究者提出 CanDO，一种测试时规范化方法，仅通过前向查询即可恢复黑盒点云模型在旋转分布偏移下损失的性能。该方法利用无导数优化搜索能最大化模型置信度的输入变换，无需模型权重或梯度。在物体部件分割任务上，CanDO 在方位角偏转下恢复 76% 的 oracle 差距，在任意 SO(3) 旋转下恢复 51%，并被 WACV 2027 接收。
正文 · 原文
Abstract:Point cloud models are increasingly deployed through score-only black-box interfaces. Under this access model, test-time rotations outside the training distribution can substantially degrade performance. Although existing mitigations such as equivariant retraining and gradient-based adaptation are effective, they are infeasible without access to model weights or gradients. To address this limitation, we propose CanDO, a test-time canonicalization method that recovers lost performance using only forward queries. Specifically, CanDO searches for an input transformation that maximizes model confidence and uses derivative-free optimization to perform this search without gradients. We further identify three conditions for successful recovery: a training augmentation gap, an accuracy-aligned confidence surrogate, and global search over a multi-modal objective. Across multiple architectures and benchmarks, CanDO recovers 76% of the oracle gap under azimuthal misorientation and 51% under arbitrary SO(3) rotations for object part segmentation. These results show that frozen score-only point cloud models can recover from rotation-induced distribution shifts through post-deployment canonicalization alone.
| Comments: | WACV 2027 |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.34850 [cs.CV] |
| (or arXiv:2610.34850v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.34850 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Seonghwan Park [view email]
[v1]
Mon, 26 Oct 2026 05:07:36 UTC (674 KB)
来源：arXiv cs.AI · arxiv.org