English

Hybrid Long and Short Range Flows for Point Cloud Filtering

Graphics 2025-08-13 v1 Computer Vision and Pattern Recognition

Abstract

Point cloud capture processes are error-prone and introduce noisy artifacts that necessitate filtering/denoising. Recent filtering methods often suffer from point clustering or noise retaining issues. In this paper, we propose Hybrid Point Cloud Filtering (HybridPF\textbf{HybridPF}) that considers both short-range and long-range filtering trajectories when removing noise. It is well established that short range scores, given by xlogp(xt)\nabla_{x}\log p(x_t), may provide the necessary displacements to move noisy points to the underlying clean surface. By contrast, long range velocity flows approximate constant displacements directed from a high noise variant patch x0x_0 towards the corresponding clean surface x1x_1. Here, noisy patches xtx_t are viewed as intermediate states between the high noise variant and the clean patches. Our intuition is that long range information from velocity flow models can guide the short range scores to align more closely with the clean points. In turn, score models generally provide a quicker convergence to the clean surface. Specifically, we devise two parallel modules, the ShortModule and LongModule, each consisting of an Encoder-Decoder pair to respectively account for short-range scores and long-range flows. We find that short-range scores, guided by long-range features, yield filtered point clouds with good point distributions and convergence near the clean surface. We design a joint loss function to simultaneously train the ShortModule and LongModule, in an end-to-end manner. Finally, we identify a key weakness in current displacement based methods, limitations on the decoder architecture, and propose a dynamic graph convolutional decoder to improve the inference process. Comprehensive experiments demonstrate that our HybridPF achieves state-of-the-art results while enabling faster inference speed.

Keywords

Cite

@article{arxiv.2508.08542,
  title  = {Hybrid Long and Short Range Flows for Point Cloud Filtering},
  author = {Dasith de Silva Edirimuni and Xuequan Lu and Ajmal Saeed Mian and Lei Wei and Gang Li and Scott Schaefer and Ying He},
  journal= {arXiv preprint arXiv:2508.08542},
  year   = {2025}
}
R2 v1 2026-07-01T04:45:23.790Z