PointNDF: Cascaded Neural Displacement Fields for Point Cloud Denoising

查看原文 / View Original 编辑 / Edit

Computer Science > Computer Vision and Pattern Recognition arXiv:2611.05411 (cs) [Submitted on 28 Oct 2026] Title:PointNDF: Cascaded Neural Displacement Fields for Point Cloud Denoising Authors:Michael Y. Wang, Pingping Cai

计算机科学 > 计算机视觉与模式识别 arXiv:2611.05411 (cs) [提交于 2026 年 10 月 28 日] 标题：PointNDF：用于点云去噪的级联神经位移场作者：Michael Y. Wang, Pingping Cai

Abstract: Point cloud denoising aims to recover clean geometry from noisy 3D observations while preserving both fine surface details and the underlying spatial distribution of the points. Many learning-based methods adopt a two-stage pipeline that first extracts point-wise features from the noisy input and subsequently predicts point-wise displacements via simple regressions.

摘要：点云去噪旨在从含噪的 3D 观测数据中恢复干净的几何结构，同时保留精细的表面细节和点云底层的空间分布。许多基于学习的方法采用两阶段流水线，首先从含噪输入中提取逐点特征，随后通过简单的回归预测逐点位移。

However, these approaches often lack robust feature guidance and distribution-aware geometric refinement, causing them to fail at maintaining optimal point uniformity and geometric quality. To address these limitations, we propose PointNDF, a denoising framework that combines patch-level geometric context with local displacement prediction.

然而，这些方法往往缺乏稳健的特征引导和具备分布感知能力的几何细化，导致它们无法保持最佳的点均匀性和几何质量。为了解决这些局限性，我们提出了 PointNDF，这是一个结合了补丁级几何上下文与局部位移预测的去噪框架。

Specifically, a Point Transformer V3 (PTv3) encoder aggregates multiscale geometric information into a global patch descriptor. Rather than directly regressing an unconstrained displacement vector, a Neural Displacement Field (NDF) fuses this global patch descriptor with per-point neighborhood features.

具体而言，Point Transformer V3 (PTv3) 编码器将多尺度几何信息聚合为全局补丁描述符。神经位移场 (NDF) 没有直接回归无约束的位移向量，而是将该全局补丁描述符与逐点邻域特征进行融合。

It then predicts each displacement through bounded-vector and scale factor branches, thereby achieving a superior, more uniform point distribution. Three cascaded NDF stages propagate local features and progressively refine the point positions. Experiments on PU-Net and ModelNet10 demonstrate competitive denoising accuracy and favorable inference efficiency.

随后，它通过有界向量分支和比例因子分支预测每个位移，从而实现更优越、更均匀的点分布。三个级联的 NDF 阶段传播局部特征并逐步细化点位置。在 PU-Net 和 ModelNet10 上的实验证明了其具有竞争力的去噪精度和良好的推理效率。