Point-Selection Fine-Tuning Framework for Robust Point Cloud Classification
Quick Answer
The PSFT framework enhances point cloud classification robustness by selectively retaining minimally influential points, achieving significant error reduction on ModelNet-C and ModelNet40-C benchmarks.
Quick Take
It outperforms existing tuning strategies, particularly with ULIP-2 and Uni3D-B models, demonstrating superior performance against corruption.
Key Points
- PSFT estimates point-wise influence to suppress outliers effectively.
- The framework uses a lightweight feature filter to refine patch-token representations.
- Extensive experiments show consistent error reduction across all tested 3D models.
- Achieved strongest results on ScanObjectNN-C with ULIP-2 and Uni3D-B.
- Implementation details available at the provided arXiv link.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Noisy and corrupted points can substantially degrade point cloud recognition performance, especially under challenging corruption settings. In particular, full fine-tuning of 3D pre-trained models may amplify the influence of outliers and overwrite robustness priors learned during pre-training, while naive parameter-efficient adaptation remains sensitive to corrupted tokens. To address this issue, we propose PSFT, a point-selection fine-tuning framework that improves robustness while remaining parameter-efficient. PSFT first estimates point-wise influence from pre-pooling features and adaptively retains minimally influential points to suppress outliers. Based on the selected subset, a prompt generation branch predicts layer-wise prompt tokens and injects them into a frozen backbone for lightweight downstream adaptation. To further mitigate residual noise after selection, we append a lightweight feature filter with bottleneck MLP transformation and Beta-gated residual blending to refine patch-token representations before prediction. Extensive experiments show that PSFT consistently reduces corruption error on ModelNet-C and ModelNet40-C across all tested 3D pre-trained backbones, while achieving the strongest ScanObjectNN-C results with ULIP-2 and Uni3D-B among the evaluated tuning strategies. Our implementation can be found at this https URL.
| Subjects: | Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2607.19711 [cs.CV] |
| (or arXiv:2607.19711v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2607.19711 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Da Li [view email]
[v1]
Wed, 22 Jul 2026 03:27:18 UTC (405 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →ProMoE-FL: Prototype-conditioned Mixture of Experts for Multimodal Federated Learning with Missing Modalities
ProMoE-FL introduces a Prototype-conditioned Mixture-of-Experts framework for multimodal federated learning, effectively addressing missing modalities. It outperforms existing methods on four chest X-ray datasets, demonstrating superior feature synthesis capabilities in both homogeneous and heterogeneous settings.