{"slug": "pointndf-cascaded-neural-displacement-fields-for-point-cloud-denoising", "title": "PointNDF: Cascaded Neural Displacement Fields for Point Cloud Denoising", "summary": "Researchers posted arXiv paper 2611.05411v1 introducing PointNDF, a point cloud denoising framework that pairs a Point Transformer V3 (PTv3) encoder with cascaded Neural Displacement Field stages. PointNDF fuses a global patch descriptor with per-point neighborhood features and predicts displacements through bounded-vector and scale factor branches across three cascaded NDF stages, which the authors say yields more uniform point distribution than two-stage pipelines that regress unconstrained displacement vectors. Experiments on PU-Net and ModelNet10 showed competitive denoising accuracy and favorable inference efficiency, with source code released at github.com/peridous/PointNDF.", "body_md": "arXiv:2611.05411v1 Announce Type: new \nAbstract: 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. 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. 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. 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. The source code is publicly available at https://github.com/peridous/PointNDF.", "url": "https://wpnews.pro/news/pointndf-cascaded-neural-displacement-fields-for-point-cloud-denoising", "canonical_source": "https://arxiv.org/abs/2611.05411", "published_at": "2026-11-05 05:00:00+00:00", "updated_at": "2026-11-05 05:23:13.612767+00:00", "lang": "en", "topics": ["computer-vision", "machine-learning", "ai-research", "neural-networks"], "entities": ["PointNDF", "Point Transformer V3", "PU-Net", "ModelNet10", "arXiv", "GitHub"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/pointndf-cascaded-neural-displacement-fields-for-point-cloud-denoising", "markdown": "https://wpnews.pro/news/pointndf-cascaded-neural-displacement-fields-for-point-cloud-denoising.md", "text": "https://wpnews.pro/news/pointndf-cascaded-neural-displacement-fields-for-point-cloud-denoising.txt", "jsonld": "https://wpnews.pro/news/pointndf-cascaded-neural-displacement-fields-for-point-cloud-denoising.jsonld"}}