3D Point Cloud Completion Method Integrating 3D Gaussian Splatting and Score Distillation Sampling

Liu Licheng; Liu Baijing; Pan Dan; Zeng An; Yang Baoyao; Zhao Jingliang

doi:10.12052/gdutxb.240133

English Version Journal Login

您当前的位置：

首页 >

文章列表页 >

3D Point Cloud Completion Method Integrating 3D Gaussian Splatting and Score Distillation Sampling

更新时间：2026-05-27

- 3D Point Cloud Completion Method Integrating 3D Gaussian Splatting and Score Distillation Sampling
- Vol. 42, Issue 4, Pages: 20-28(2025)
- 作者机构：
  
  1. 广东工业大学信息工程学院,广东,广州,510006
  2. 广东技术师范大学电子与信息学院,广东,广州,510665
  3. 广东工业大学计算机学院,广东,广州,510006
- 作者简介：
- 基金信息：
- DOI：10.12052/gdutxb.240133
  CLC： TP183
- Published：2025
- 稿件说明：
移动端阅览
刘立程, 刘柏菁, 潘丹, 等. 融合三维高斯点染和分数蒸馏采样的三维点云补全方法[J]. 广东工业大学学报, 2025, 42(4): 20–28.

Liu Licheng, Liu Baijing, Pan Dan, et al. 3D Point Cloud Completion Method Integrating 3D Gaussian Splatting and Score Distillation Sampling[J]. 2025, 42(4): 20-28.
刘立程, 刘柏菁, 潘丹, 等. 融合三维高斯点染和分数蒸馏采样的三维点云补全方法[J]. 广东工业大学学报, 2025, 42(4): 20–28. DOI： 10.12052/gdutxb.240133

Liu Licheng, Liu Baijing, Pan Dan, et al. 3D Point Cloud Completion Method Integrating 3D Gaussian Splatting and Score Distillation Sampling[J]. 2025, 42(4): 20-28. DOI： 10.12052/gdutxb.240133.

摘要

针对现有三维点云补全方法依赖于配对数据的有监督学习，导致数据成本高昂且泛化能力不足的问题，本文提出了一种无需配对数据进行训练的三维点云补全方法。结合三维高斯点染技术和分数蒸馏采样技术，将不完整点云转换为三维高斯模型，并通过预训练的二维扩散模型引导三维高斯模型逐步优化，预测完整点云，实现三维点云补全任务。引入点密集化操作以增加输入点云的稠密度，增强初始三维高斯的质量，并采用渐进式相机采样策略在优化初期控制相机采样范围，提升了优化效率。实验结果表明，本文方法在RedWood-3DScan数据集上取得了优于现有方法的性能。消融实验进一步验证了所提出优化方案的有效性，证明了本方法在处理真实数据时的优越性。

Abstract

This study addressed the limitations of existing 3D point cloud completion methods

which rely on paired data and supervised learning

resulting in high data costs and limited generalization capabilities. This paper proposed a novel 3D point cloud completion method without requiring paired data for training by integrating 3D Gaussian splatting and Score Distillation Sampling (SDS) techniques. In the proposed method

incomplete point clouds were transformed into 3D Gaussian models

which were iteratively optimized using a pre-trained 2D diffusion model to predict complete point clouds and fulfill the completion task. To enhance the initial 3D Gaussian models

the authors introduced a point densification technique that increased the density of input point clouds. Additionally

a progressive camera sampling strategy was adopted during the early optimization stages to control the camera sampling range

thereby improving optimization efficiency. The experimental results demonstrate that the proposed method outperforms existing approaches on the RedWood-3DScan dataset. Ablation studies further validate the effectiveness of the optimization strategies

confirming the superiority of the proposed method in handling real-world data.

关键词

Keywords

references

YUAN W, KHOT T, HELD D, et al . Pcn: point completion network[C ] //2018 International Conference on 3D vision (3DV) . Verona: IEEE, 2018: 728-737.

REN J, PAN L, LIU Z. Benchmarking and analyzing point cloud classification under corruptions[C] Proceedings of the 39th International Conference on Machine Learning (ICML) . Baltimore: IMLS, 2022, 162: 18559-18575.

ZHU J J, LIU D F. High-resolution feature network for 3d point cloud segmentation and classification[J]. Journal of Guangdong University of Technology, 2025, 42(3): 44-51.

NGUYEN T A, LEE S. 3D orientation and object classification from partial model point cloud based on pointnet[C]//2018 IEEE International Conference on Image Processing, Applications and Systems (IPAS) . Sophia Antipolis: IEEE, 2018: 192-197.

ZHANG J M, CHEN W J, WANG Y P, et al . Point set voting for partial point cloud analysis[J ] . IEEE Robotics and Automation Letters (RA-L) , 2021, 6(2): 596-603.

ZHUANG Z Y, ZHI Z Y, HAN T, et al . A survey of point cloud completion[J ] . IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2024, 17: 5691-5 711.

WANG X G, ANG J M H, LEE G H. Cascaded refinement network for point cloud completion[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . Seattle: IEEE, 2020: 790-799.

SARMAD M, LEE H J, KIM Y M. Rl-gan-net: a reinforcement learning agent controlled gan network for real-time point cloud shape completion[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . Long Beach: IEEE, 2019: 5898-5907.

WU X Z, WU X F, LUAN T Y, et al . FSC: few-point shape completion[C ] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . Seattle: IEEE, 2024: 26077-26087.

WANG J, CUI Y, GUO D Y, et al . Pointattn: you only need attention for point cloud completion[C ] //Proceedings of the AAAI Conference on Artificial Intelligence (AAAI) . Vancouver: AAAI, 2024, 38(6): 5472-5480.

ZHAO L D, HU Y H, YANG X, et al . Robust multi-task learning network for complex LiDAR point cloud data preprocessing[J ] . Expert Systems with Applications, 2024, 237: 121552.

ACHITUVE I, MARON H, CHECHIK G. Self-supervised learning for domain adaptation on point clouds[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) . Waikoloa: IEEE, 2021: 123-133.

ZHANG J Z, CHEN X Y, CAI Z G, et al . Unsupervised 3D shape completion through gan inversion[C ] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . Nashville: IEEE, 2021: 1768-1777.

XIAO H H, LI Y Q, KANG W X, et al . Distinguishing and matching-aware unsupervised point cloud completion[J ] . IEEE Transactions on Circuits and Systems for Video Technology, 2023, 33(9): 5160-5173.

CAI Y J, LIN K Y, ZHANG C, et al . Learning a structured latent space for unsupervised point cloud completion[C ] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . New Orleans: IEEE, 2022: 5543-5553.

ROMBACH R, BLATTMANN A, LORENZ D, et al . High-resolution image synthesis with latent diffusion models[C ] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . New Orleans: IEEE, 2022: 10684-10695.

KASTEN Y, RAHAMIM O, CHECHIK G. Point cloud completion with pretrained text-to-image diffusion models[C]//Advances in Neural Information Processing Systems (NeurIPS). New Orleans: Curran Associates, 2023, 36: 12171-12191.

KERBL B, KOPANAS G, LEIMKÜHLER T, et al . 3D gaussian splatting for real-time radiance field rendering[J ] . ACM Transactions on Graphics (TOG) , 2023, 42(4): 1-14.

MILDENHALL B, SRINIVASAN P P, TANCIK M, et al . Nerf: representing scenes as neural radiance fields for view synthesis[J ] . Communications of the ACM, 2021, 65(1): 99-106.

YI T R, FANG J M, WANG J J, et al . GaussianDreamer: fast generation from text to 3D gaussians by bridging 2D and 3D diffusion models[C ] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . Seattle: IEEE, 2024: 6796-6807.

BARRON J T, MILDENHALL B, TANCIK M, et al . Mip-NeRF: a multiscale representation for anti-aliasing neural radiance fields[C ] //Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) . Montreal: IEEE, 2021: 5855-5864.

ZHANG X C, FENG Y T, LI S Q, et al . View-guided point cloud completion[C ] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . Nashville: IEEE, 2021: 15890-15899.

WANG Y F, FELICE S, WU S H, et al . Differentiable surface splatting for point-based geometry processing[J ] . ACM Transactions on Graphics (TOG) , 2019, 38(6): 1-14.

KOPANAS G, LEIMKÜHLER T, RAINER G, et al . Neural point catacaustics for novel-view synthesis of reflections[J ] . ACM Transactions on Graphics (TOG) , 2022, 41(6): 1-15.

PASZKE A, GROSS S, MASSA F, et al . Pytorch: an imperative style, high-performance deep learning library[C ] //Advances in Neural Information Processing Systems (NeurIPS) , Vancouver: Curran Associates, 2019, 32: 8026-8037.

PAN L, CHEN X Y, CAI Z A, et al . Variational relational point completion network[C ] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . Nashville: IEEE, 2021: 8524-8533.

ZENG A, LUO L, PAN D, et al . Multi-task segmentation network for the plant on 3D point cloud[J ] . Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE) , 2023, 39(12): 132-140.

YU X M, RAO Y M, WANG Z Y, et al . Pointr: diver se point cloud completion with geometry-aware transformers[C ] //Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) . Montreal: IEEE, 2021: 12498-12507.

YAN X G, LIN L Q, MITRA N J, et al . Shapeformer: transformer-based shape completion via sparse representation[C ] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . New Orleans: IEEE, 2022: 6239-6249.

WU R D, CHEN X L, ZHUANG Y X, et al . Multimodal shape completion via conditional generative adversarial networks[C ] //Computer Vision – ECCV 2020. Glasgow: Springer, 2020: 281-296.

HONG S M, YAVARTANOO M, NESHATAVAR R, et al . ACL-SPC: adaptive closed-loop system for self-supervised point cloud completion[C ] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . Vancouver: IEEE, 2023: 9435-9444.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Robust Design for Artificial Noise-assisted UAV Terahertz Secure Communication

Efficient Temporal Modeling for RGB-T Tracking

SCA-YOLO: Real-time Target Detection Based on Deformable Convolution and Context-aware Attention

Energy Management Strategy of Hybrid Locomotive Based on Road Condition Parameters

Multi-Scale Contrastive Embedding Enhanced Adaptive Intrusion Detection Model

Related Author

No data

Related Institution

No data

AI问答

⁰