Lune

CVPR Findings2026

PlanGS: Active 3D Gaussian Reconstruction with Real-Time Planning

Wenxiang Xie, Anpei Chen, Haoming Yu, Yujun Shen, Weiwei Xu

Abstract

Recent advances in 3D Gaussian Splatting (3DGS) have made it a mainstream representation for novel view synthesis and 3D reconstruction, owing to its high rendering efficiency and reconstruction quality. However, existing approaches typically rely on manually collected image data, which suffers from low acquisition efficiency and limited scene coverage. To overcome these limitations, we propose an active reconstruction framework that enables an autonomous agent to efficiently explore unknown environments and collect informative observations for highquality reconstruction. Our method employs Poisson-disk sampling to generate uniformly distributed candidate viewpoints and constructs a path graph structure for efficient exploration. In addition, we propose an A*-based adaptive planner that dynamically balances local exploration and global guidance. Extensive experiments demonstrate that our approach achieves state-of-the-art photometric accuracy, while achieving the shortest exploration time among comparable methods. Furthermore, we validate the feasibility of our system on a quadruped robot, showcasing its potential for real-world autonomous 3D reconstruction.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

  • 3D Gaussian Splatting for Real-Time Radiance Field Rendering cites Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance Fields
Builds onCited byLeft to right by year; the more cited, the higher.

Builds on23

  1. 3D Gaussian Splatting for Real-Time Radiance Field RenderingBernhard Kerbl, Georgios Kopanas, Thomas Leimkühler, George DrettakisSIGGRAPH 2023 · 5,687 citations
  2. Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance FieldsJonathan T. Barron, Ben Mildenhall, Matthew Tancik, Peter Hedman et al.ICCV 2021 · 2,700 citations
  3. NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view ReconstructionPeng Wang, Lingjie Liu, Yuan Liu, Christian Theobalt et al.NeurIPS 2021 · 2,500 citations
  4. Habitat: A Platform for Embodied AI ResearchManolis Savva, Jitendra Malik, Devi Parikh, Dhruv Batra et al.ICCV 2019 · 1,863 citations
  5. Volume Rendering of Neural Implicit SurfacesLior Yariv, Jiatao Gu, Yoni Kasten, Yaron LipmanNeurIPS 2021 · 1,421 citations

Related papers

  1. ActiveGAMER: Active GAussian Mapping through Efficient RenderingLiyan Chen, Huangying Zhan, Kevin Chen, Xiangyu Xu et al.CVPR 2025
  2. Understanding while Exploring: Semantics-driven Active MappingLiyan Chen, Huangying Zhan, Hairong Yin, Yi Xu et al.NeurIPS 2025 · 5 citations
  3. DG-SLAM: Robust Dynamic Gaussian Splatting SLAM with Hybrid Pose OptimizationYueming Xu, Haochen Jiang, Zhongyang Xiao, Jianfeng Feng et al.NeurIPS 2024 · 65 citations
  4. ExploreGS: Explorable 3D Scene Reconstruction with Virtual Camera Samplings and Diffusion PriorsMinsu Kim, Subin Jeon, In Cho, Mijin Yoo et al.ICCV 2025 · 1 citation
  5. Path Matters: Unveiling Geometric Implicit Bias via Curvature-Aware Sparse View OptimizationCanran Xiao, Liaoyuan Fan, Yanbin Li, Jing Tang et al.ICLR 2026
Dusk over the sea between two cliffs drawn in fine vertical lines