3D Gaussian Flats: Hybrid 2D/3D Photometric Scene Reconstruction
Not on the map. Why?
Abstract (arXiv)
Recent advances in radiance fields and novel view synthesis enable creation of realistic digital twins from photographs. However, current methods struggle with flat, texture-less surfaces, creating uneven and semi-transparent reconstructions, due to an ill-conditioned photometric reconstruction objective. Surface reconstruction methods solve this issue but sacrifice visual quality. We propose a novel hybrid 2D/3D representation that jointly optimizes constrained planar (2D) Gaussians for modeling flat surfaces and freeform (3D) Gaussians for the rest of the scene. Our end-to-end approach dynamically detects and refines planar regions, improving both visual fidelity and geometric accuracy. It achieves state-of-the-art depth estimation on ScanNet++ and ScanNetv2, and excels at mesh extraction without overfitting to a specific camera model, showing its effectiveness in producing high-quality reconstruction of indoor scenes.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 32 of its 38 references that are in the map. First posted on arXiv 19 Sep 2025.
Builds on
32 of 38 references.
References · 32
- 20233D Gaussian Splatting for Real-Time Radiance Field Renderingcited by 6,226
- 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,117
- 2024SAM 2: Segment Anything in Images and Videoscited by 2,863
- 2021Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fieldscited by 2,032
- 2021NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstructioncited by 1,712
- 2024Depth Anything: Unleashing the Power of Large-Scale Unlabeled Datacited by 1,360
- 20242D Gaussian Splatting for Geometrically Accurate Radiance Fieldscited by 1,076
- 2023ScanNet++: A High-Fidelity Dataset of 3D Indoor Scenescited by 764
- 2023SuGaR: Surface-Aligned Gaussian Splatting for Efficient 3D Mesh Reconstruction and High-Quality Mesh Renderingcited by 616
- 2021RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from Sparse Inputscited by 615
- 2023Neuralangelo: High-Fidelity Neural Surface Reconstructioncited by 467
- 2021Advances in Neural Renderingcited by 365
20 more
- 2024PGSR: Planar-based Gaussian Splatting for Efficient and High-Fidelity Surface Reconstructioncited by 285
- 2024Gaussian Opacity Fields: Efficient Adaptive Surface Reconstruction in Unbounded Scenescited by 232
- 2024A Survey on 3D Gaussian Splattingcited by 224
- 20243D Gaussian Splatting as Markov Chain Monte Carlocited by 217
- 2024DN-Splatter: Depth and Normal Priors for Gaussian Splatting and Meshingcited by 138
- 2022Neural 3D Scene Reconstruction with the Manhattan-world Assumptioncited by 123
- 2024RaDe-GS: Rasterizing Depth in Gaussian Splattingcited by 111
- 2018PlaneRCNN: 3D Plane Detection and Reconstruction from a Single Imagecited by 105
- 2023Bayes' Rays: Uncertainty Quantification for Neural Radiance Fieldscited by 91
- 2023Adaptive Shells for Efficient Neural Radiance Field Renderingcited by 69
- 2022StructNeRF: Neural Radiance Fields for Indoor Scenes with Structural Hintscited by 29
- 2022PlanarRecon: Real-time 3D Plane Detection and Reconstruction from Posed Monocular Videoscited by 27
- 2022NeurMiPs: Neural Mixture of Planar Experts for View Synthesiscited by 23
- 2023Fast View Synthesis of Casual Videos with Soup-of-Planescited by 22
- 2024AirPlanes: Accurate Plane Estimation via 3D-Consistent Embeddingscited by 13
- 2024PlanarSplatting: Accurate Planar Surface Reconstruction in 3 Minutescited by 13
- 2021PlaneRecNet: Multi-Task Learning with Cross-Task Consistency for Piece-Wise Plane Detection and Reconstruction from a Single RGB Imagecited by 8
- 2024Planar Gaussian Splattingcited by 6
- 2023PlanarNeRF: Online Learning of Planar Primitives with Neural Radiance Fieldscited by 5
- 2024Integrating Meshes and 3D Gaussians for Indoor Scene Reconstruction with SAM Mask Guidancecited by 4
Built on by
Citing papers per year: 2026: 2.
2 papers.