Pith. sign in

REVIEW 3 major objections 5 minor 52 references

CityGo: Lightweight Urban Modeling and Rendering with Proxy Buildings and Residual Gaussians

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read CityGo combines textured proxy buildings with residual and surrounding 3D Gaussians to make large-scale aerial urban scenes render in real time on mobile GPUs, cutting training time by about 1.4x compared with pure 3D Gaussian Splatting…

desk verdict A promising hybrid mesh-Gaussian urban renderer with real efficiency gains, but the central speedup claim is confounded by an unmatched SH order and needs a control run. read the letter →

arxiv 2505.21041 v3 pith:DDPZAMYJ submitted 2025-05-27 cs.GR cs.CV

Weihang Liu , Yuhui Zhong , Yuke Li , Xi Chen , Jiadi Cui , Honglong Zhang , Lan Xu , Xin Lou
show 3 more authors
Yujiao Shi Jingyi Yu Yingliang Zhang
This is my paper
classification cs.GRcs.CV
keywords hybridscenerepresentation3DGaussianSplattingproxygeometryurbanmodelingaerialimageryreal-timerenderingmobileGPUtexturemapping
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

CityGo claims that large-scale urban scenes from aerial imagery can be modeled far more cheaply than pure 3D Gaussian Splatting allows, by splitting the scene into two parts: buildings become lightweight textured proxy meshes, while everything else stays as sparse 3D Gaussians. The paper shows that this hybrid representation trains about 1.4 times faster than standard 3DGS, uses dramatically less memory, and still renders 1.5 km² scenes at a minimum of 20 FPS on a Jetson AGX Orin mobile GPU. If true, this makes photorealistic city-scale digital twins, AR navigation, and UAV inspection feasible on edge devices rather than only on high-end desktop GPUs.

What carries the argument

The load-bearing mechanism is the three-part hybrid representation. Building proxy meshes are extracted from MVS point clouds through a bottom-up, cluster-aware version of layer-based proxy reconstruction (Building Point Cloud Completion), then refined by planar primitive fitting and kinetic shape reconstruction; textures are generated from zero-order SH Gaussian renderings via the TwinTex method and refined with differentiable UV optimization. Residual Gaussians are selected by a color residual map score, which measures where proxy-rendered color disagrees with ground truth photos, and their placement is constrained near the mesh surface using a depth guard interval. Surrounding Gaussians are kept from the original full-scene Gaussians but importance-sampled (following Mini-Splatting) to remove redundancy. Together, these components make the representation both compact enough for mobile inference and fine-grained enough to recover facade details.

What would settle it

Run CityGo on a dense urban scene containing a large non-building structure attached to a building (e.g., a tower crane or a large sign). If the segmentation mistakenly treats that structure as part of the building, the fixed-opacity proxy texture will bake its color onto the facade, and held-out-view PSNR should drop by substantially more than the roughly 1 dB gap CityGo already accepts versus 3DGS—an observation that would confirm the stated limitation and show that the hybrid representation loses its fidelity advantage exactly when proxy geometry fails.

Watch

Extended reading notes

Core claim

The central claim is that a hybrid scene representation—compact textured proxy geometry for buildings, sparse residual Gaussians for high-frequency building details, and importance-downsampled surrounding Gaussians for unstructured environment—recovers the visual quality of dense 3D Gaussian Splatting at a fraction of the cost. The authors demonstrate on aerial datasets and the UrbanBIS benchmark that their method achieves a 1.4x average training speedup, reduces model size to roughly 1/8 of 3DGS on their own captures, and sustains real-time rendering (161–196 FPS on an RTX A6000, 20–51 FPS on a mobile GPU) with PSNR within about 1 dB of 3DGS. The key insight is that buildings, which dominate city scenes, can be represented as stable planar-proxy meshes with baked textures, leaving only the subtle, view-dependent, and geometrically complex parts to be carried by Gaussians.

Load-bearing premise

The whole pipeline assumes that the proxy-building extraction—especially the GIS building masks and the MVS point cloud completion—produces geometry accurate enough that the fixed-opacity baked textures do not need to be corrected by the residual Gaussians.

Editorial extensions

If this is right

  • If CityGo's claims hold, city-scale digital twins and AR navigation layers can be deployed on smartphones, drones, and in-vehicle hardware without a cloud backend, since rendering is 20+ FPS on a mobile GPU and training runs on a single desktop GPU.
  • The 1.4x training speedup and roughly 1/8 model-size reduction relative to 3DGS mean that urban scenes can be reconstructed and updated frequently, making practical the kind of 'living city' models that would otherwise require costly re-optimization.
  • The hybrid representation decouples structured buildings, which can be edited or replaced, from unstructured surroundings, so urban-planning tools could swap in new building designs while keeping the Gaussian environment intact.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same hybrid logic could extend beyond buildings: any class with predictable planar structure (e.g., roads, bridges, parking structures) could be absorbed into proxy geometry, potentially shrinking the Gaussian budget even further on infrastructure-heavy scenes.
  • Because the color residual map explicitly exposes where the proxy fails, CityGo's component selection is effectively an automatic budget allocator: it spends Gaussian capacity exactly where the mesh cannot reproduce the photograph, which suggests it could be tuned against a perceptual loss rather than PSNR to yield even smaller footprints.
  • A testable extension would be to replace the fixed proxy opacity of 1.0 with an optimizable transparency, which would let the optimization stage correct misclassified non-building structures instead of baking their colors into the facade—an evolution the paper's own limitation section points toward.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. CityGo proposes a hybrid representation for large-scale aerial urban scenes that combines textured proxy meshes for buildings with two types of 3D Gaussians: residual Gaussians that capture high-frequency details on building surfaces and downsampled surrounding Gaussians for the rest of the environment. The pipeline starts from MVS point clouds, completes building geometry via a layer-based point-cloud completion method (BPCC), extracts planar proxy meshes via FPP/KSR, generates occlusion-free textures using zero-order SH Gaussians and TwinTex-style rendering, and then jointly optimizes proxy textures and Gaussian parameters. The paper reports a roughly 1.4x training-time speedup over pure 3DGS, comparable PSNR, large reductions in model size and Gaussian count, and real-time rendering on a Jetson AGX Orin.

Significance. If the reported efficiency and fidelity results hold, CityGo would be a practically useful step toward city-scale photorealistic rendering on edge devices, and the hybrid mesh-plus-residual-Gaussian representation is a sensible design direction for aerial urban data. The paper has concrete strengths: it evaluates on two large drone-captured scenes (1.5 km² each) plus the UrbanBIS benchmark, reports mobile GPU FPS, and quantifies the proxy-mesh compression (61.7x fewer vertices/faces). The ablation on UV finetuning and CRM-based sampling is useful, and the hybrid rendering formulation in Eq. (4) is explicit. However, the central quantitative claims rest on comparisons in which the SH order is not matched between CityGo and the Gaussian baselines, which makes the headline speedup and 'comparable fidelity' statements not yet established.

major comments (3)
  1. [Section 5.2, Tables 1-2, Section 5.1] The main comparison is confounded by an unmatched SH order. The paper states in Section 5.1 that CityGo uses SH=0 'to accelerate convergence,' while Section 5.2 reports that the 3DGS and 2DGS baselines are trained with SH=2. SH=0 removes view-dependent appearance and reduces the per-Gaussian parameter count and per-pixel evaluation cost, which directly lowers both training time and memory. The 1.14-1.4x training speedup and the 'comparable fidelity' claim can therefore be attributed partly or entirely to the cheaper appearance model rather than to the hybrid proxy-plus-residual representation. This is an internal control gap, and it is testable. I strongly recommend adding an SH-matched ablation: either 3DGS with SH=0 (and 2DGS with SH=0) or a CityGo variant with SH=2, reporting PSNR, training time, model size, and FPS. Without this, the central claim of a 1.4x speedup from the hybrid representation is not supported.
  2. [Section 5.2, 'Time (h)' columns] It is unclear what is included in the reported training time for CityGo. The pipeline described in Sections 3-4 has multiple stages: initial SH=0 3DGS training, BPCC point-cloud completion, FPP/KSR mesh extraction, synthetic texture rendering via TwinTex, UV finetuning, residual Gaussian selection and 100K-iteration optimization, and surrounding Gaussian downsampling with 30K-iteration optimization. If 'Time' for CityGo counts only the Gaussian optimization stages, while the time for 3DGS counts the full end-to-end training, the speedup comparison is not apples-to-apples. Please specify the timing methodology, list the per-stage timings, and state whether mesh extraction and texturing are included in the reported 17.6 h / 16.6 h / 1.4 h figures.
  3. [Abstract, Tables 1-2] The abstract's claim of 'on average 1.4x speedup' is not fully supported by the tables. The speedups implied by the reported training times are 1.6/1.4 = 1.14x on UrbanBIS (Table 1), 24.3/17.6 = 1.38x on Area-H, and 23.4/16.6 = 1.41x on Area-L (Table 2). The average across all three datasets is approximately 1.31x, not 1.4x. If the '1.4x' is intended to describe only the two aerial datasets, that should be stated explicitly; otherwise the abstract's quantitative claim should be corrected to match the reported data.
minor comments (5)
  1. [Section 3.1] The BPCC thresholds gamma (area ratio for introducing a new dominant contour) and beta (hole-filling ratio) are described as fixed, but no values are given and no sensitivity analysis is provided; please report the chosen values and, ideally, an ablation or robustness check.
  2. [Section 3.3, Eq. (5)] In Eq. (5), the notation P_gamma for the set of rays from 'viewpoint-gamma' is inconsistent with the later use of gamma as a threshold; please rename one of these to avoid confusion.
  3. [Section 5.2] The FPS numbers in Table 2 are measured at 1988x1326 on an A6000, while Table 3 uses 720p on a Jetson AGX Orin; the caption of Table 3 states the resolution, but the main text does not draw attention to this resolution difference when discussing real-time performance.
  4. [Section 5.3, Table 4] The ablation table reports PSNR and size but no error bars or multiple runs; although the differences are small, stating how many runs were averaged (if any) would strengthen the ablation.
  5. [Throughout] There are several typos and grammatical slips, e.g., 'UA Vs' in the Introduction, 'comsumption' in Section 5.2, 'based using CRM' in Section 3.3, and inconsistent capitalization 'CityGO' in the captions of Figures 7 and 8.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: CityGo's central claims are empirical and self-contained; the SH-order mismatch is a control concern, not circular reasoning.

full rationale

No circular step found. CityGo's central claims are empirical measurements against held-out test views and external baselines (Tables 1-3), not derivations from assumptions. The hybrid representation is defined structurally as proxy meshes plus residual and surrounding Gaussians, and the claimed 1.4x training speedup is read directly from Table 2, not forced by construction. The CRM-based residual Gaussian placement uses ground-truth image colors as supervision; this is a fitting and selection mechanism, not a prediction equivalent to its input. The UV texture refinement also optimizes against ground-truth imagery, again not circular. Self-citations (LetsGo [9], TensoRF [5]) appear only in related work and do not supply any load-bearing theorem or forced choice. The SH-order mismatch (CityGo SH=0 vs 3DGS SH=2) is a legitimate experimental-control concern that may affect attribution of the speedup, but it is not a circularity: the paper discloses the SH setting, and the speedup is an empirical outcome rather than a quantity defined in terms of itself. No equation in the paper equates an output with its input, and no fitted parameter is renamed as a prediction.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The central claim relies on a chain of existing components (MVS, alpha-shape, FPP, KSR, TwinTex, 3DGS) and on several hand-set thresholds that are not publicly documented. No new physical entities are introduced.

free parameters (6)
  • residual threshold E_k = 0.2
    Gaussians with CRM score above this threshold are selected as residual Gaussians (Sec. 5.1). The choice is not justified by a sensitivity analysis.
  • BPCC area ratio threshold gamma = not specified
    In Sec. 3.1, a new dominant contour is added when Area(C)/Area(S) <= gamma; gamma is called a fixed threshold but no value or tuning procedure is given.
  • BPCC hole-filling ratio beta = not specified
    In Appendix A, points are sampled on the polyhedral cell surface when Area(F)/Area(C) < beta; beta is not specified.
  • guard interval d_g = not specified
    In Sec. 3.3, Gaussians with depth < d_m + d_g are activated; d_g controls residual Gaussian placement relative to the mesh and is never given a value.
  • surrounding Gaussian downsample factor = 0.1
    In Sec. 4 and Sec. 5.1, the initial set is downsampled by a factor of 0.1 via importance sampling; this factor is fixed without ablation across values.
  • number of vertical layers L = not specified
    BPCC slices the point cloud into L layers; L affects contour granularity and is not reported.
assumptions (6)
  • domain assumption MVS point clouds produced by Agisoft are sufficiently dense and accurate for proxy extraction.
    The entire BPCC and FPP/KSR pipeline depends on MVS output quality; the paper uses off-the-shelf [1] and does not quantify sensitivity to MVS errors.
  • domain assumption GIS-derived building segmentation reliably separates buildings from surrounding structures.
    Buildings are segmented using GIS masks before proxy extraction; misclassification of cranes or signage is explicitly admitted as a limitation in the conclusion.
  • standard math The alpha-shape and DBSCAN based contour extraction preserves the dominant structure of buildings.
    These are standard computational geometry tools; the paper relies on them to derive layered proxy geometry without formal guarantee of correctness.
  • domain assumption FPP planar primitive fitting and KSR produce a faithful proxy geometry from completed point clouds.
    The mesh extraction step uses [49] and [2] as black boxes; the accuracy of the proxy is a precondition for the depth-based guard interval and texture mapping.
  • ad hoc to paper Hybrid rendering formula (4) treats the mesh as the final surface with opacity 1.0, neglecting semi-transparent materials and view-dependent effects.
    The compositing model is introduced in Sec. 3.3; it assumes Gaussians only contribute before the mesh surface, which limits material modeling.
  • domain assumption Importance-based downsampling from Mini-Splatting preserves visual quality when reducing surrounding Gaussians by a factor of 0.1.
    The method borrows the importance metric from [12] without presenting an independent validation of the 0.1 sampling factor.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CityGo: Lightweight Urban Modeling and Rendering with Proxy Buildings and Residual Gaussians." pith.science (2026). https://pith.science/paper/DDPZAMYJ

@misc{pith2026250521041,
  author       = {Pith},
  title        = {Pith review of: CityGo: Lightweight Urban Modeling and Rendering with Proxy Buildings and Residual Gaussians},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DDPZAMYJ}},
  note         = {Machine review of arXiv:2505.21041}
}
read the original abstract

Accurate and efficient modeling of large-scale urban scenes is critical for applications such as AR navigation, UAV based inspection, and smart city digital twins. While aerial imagery offers broad coverage and complements limitations of ground-based data, reconstructing city-scale environments from such views remains challenging due to occlusions, incomplete geometry, and high memory demands. Recent advances like 3D Gaussian Splatting (3DGS) improve scalability and visual quality but remain limited by dense primitive usage, long training times, and poor suit ability for edge devices. We propose CityGo, a hybrid framework that combines textured proxy geometry with residual and surrounding 3D Gaussians for lightweight, photorealistic rendering of urban scenes from aerial perspectives. Our approach first extracts compact building proxy meshes from MVS point clouds, then uses zero order SH Gaussians to generate occlusion-free textures via image-based rendering and back-projection. To capture high-frequency details, we introduce residual Gaussians placed based on proxy-photo discrepancies and guided by depth priors. Broader urban context is represented by surrounding Gaussians, with importance-aware downsampling applied to non-critical regions to reduce redundancy. A tailored optimization strategy jointly refines proxy textures and Gaussian parameters, enabling real-time rendering of complex urban scenes on mobile GPUs with significantly reduced training and memory requirements. Extensive experiments on real-world aerial datasets demonstrate that our hybrid representation significantly reduces training time, achieving on average 1.4x speedup, while delivering comparable visual fidelity to pure 3D Gaussian Splatting approaches. Furthermore, CityGo enables real-time rendering of large-scale urban scenes on mobile consumer GPUs, with substantially reduced memory usage and energy consumption.

Figures

Figures reproduced from arXiv: 2505.21041 by the authors.

Figure 1
Figure 1. We present CityGo, an explicit and efficient framework for high-fidelity rendering of large-scale urban scenes. By combining [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of our hybrid representation for large-scale urban scenes. We begin by generating dense point clouds from aerial [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Textured proxy buildings Reconstruction. We first ap [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Qualitative comparison of our method and baseline methods on the Area-H and Area-L datasets. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Ablation studies on UV finetuning and CRM-based [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Practical Use Cases of CityGo in Urban Planning. [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: Visualization of examples rendered using our CityGO models in the Area-H scene. Comparisons are provided to highlight the [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]
Figure 8
Figure 8. Figure 8: Visualization of examples rendered using our CityGO models in the Area-L scene. Comparisons are provided to highlight the [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

52 extracted references · 45 canonical work pages

  1. [20]

    Fast Building Instance Proxy Reconstruction for Large Urban Scenes.IEEE Trans

    Jianwei Guo, Haobo Qin, Yinchang Zhou, Xin Chen, Lian- gliang Nan, and Hui Huang. Fast Building Instance Proxy Reconstruction for Large Urban Scenes.IEEE Trans. Pat- tern Anal. Mach. Intell., 46(11):7267–7282, 2024. 3, 4

    work page 2024

  2. [1]

    Agisoft photoscan user manual: Professional edi- tion.https://www.agisoft.com, 2016

    Agisoft. Agisoft photoscan user manual: Professional edi- tion.https://www.agisoft.com, 2016. Accessed: 2025-05-24. 3, 6

    work page 2016

  3. [2]

    Kinetic Shape Reconstruction.ACM Trans

    Jean-Philippe Bauchet and Florent Lafarge. Kinetic Shape Reconstruction.ACM Trans. Graph., 39(5), 2020. 5

    work page 2020

  4. [3]

    Large scale 3d morphable models.International Journal of Computer Vision, 126(2): 233–254, 2018

    James Booth, Anastasios Roussos, Allan Ponniah, David Dunaway, and Stefanos Zafeiriou. Large scale 3d morphable models.International Journal of Computer Vision, 126(2): 233–254, 2018. 3

    work page 2018

  5. [4]

    Hybrid Explicit Representation for Ultra-Realistic Head Avatars

    Hongrui Cai, Yuting Xiao, Xuan Wang, Jiafei Li, Yudong Guo, Yanbo Fan, Shenghua Gao, and Juyong Zhang. Hybrid Explicit Representation for Ultra-Realistic Head Avatars. arXiv preprint arXiv:2403.11453, 2024. 3

    arXiv 2024

  6. [5]

    TensoRF: Tensorial Radiance Fields

    Anpei Chen, Zexiang Xu, Andreas Geiger, Jingyi Yu, and Hao Su. TensoRF: Tensorial Radiance Fields. InProc. Eur. Conf. Comput. Vis., pages 333–350, 2022. 3

    work page 2022

  7. [6]

    DOGS: Distributed-Oriented Gaussian Splatting for Large-Scale 3D Reconstruction Via Gaussian Consensus

    Yu Chen and Gim Hee Lee. DOGS: Distributed-Oriented Gaussian Splatting for Large-Scale 3D Reconstruction Via Gaussian Consensus. InarXiv, 2024. 2, 3

    work page 2024

  8. [7]

    Meshgs: Adaptive mesh-aligned gaus- sian splatting for high-quality rendering

    Jaehoon Choi, Yonghan Lee, Hyungtae Lee, Heesung Kwon, and Dinesh Manocha. Meshgs: Adaptive mesh-aligned gaus- sian splatting for high-quality rendering. InProceedings of the Asian Conference on Computer Vision, pages 3310– 3326, 2024. 3

    work page 2024

Show all 52 references
  1. [8]

    Variational shape approximation.ACM Trans

    David Cohen-Steiner, Pierre Alliez, and Mathieu Desbrun. Variational shape approximation.ACM Trans. Graph., 23 (3), 2004. 3

    2004

  2. [9]

    LetsGo: Large-Scale Garage Modeling and Rendering via LiDAR-Assisted Gaussian Primitives.ACM Trans

    Jiadi Cui, Junming Cao, Fuqiang Zhao, Zhipeng He, Yifan Chen, Yuhui Zhong, Lan Xu, Yujiao Shi, Yingliang Zhang, and Jingyi Yu. LetsGo: Large-Scale Garage Modeling and Rendering via LiDAR-Assisted Gaussian Primitives.ACM Trans. Graph., 43(6), 2024. 2, 3

    2024

  3. [10]

    Edelsbrunner, D

    H. Edelsbrunner, D. Kirkpatrick, and R. Seidel. On the shape of a set of points in the plane.IEEE Transactions on Infor- mation Theory, 29(4):551–559, 1983. 4

    1983

  4. [11]

    A density-based algorithm for discovering clusters in large spatial databases with noise

    Martin Ester, Hans-Peter Kriegel, J ¨org Sander, and Xiaowei Xu. A density-based algorithm for discovering clusters in large spatial databases with noise. InProc. AAAI Conf. Artif. Intell., page 226–231, 1996. 4

    1996

  5. [12]

    Mini-Splatting: Represent- ing Scenes with a Constrained Number of Gaussians

    Guangchi Fang and Bing Wang. Mini-Splatting: Represent- ing Scenes with a Constrained Number of Gaussians. In Proc. Eur. Conf. Comput. Vis., pages 165–181, 2024. 3, 6

    2024

  6. [13]

    FlashGS: Efficient 3D Gaussian Splatting for Large-scale and High-resolution Ren- dering.arXiv preprint arXiv:2408.07967, 2024

    Guofeng Feng, Siyan Chen, Rong Fu, Zimu Liao, Yi Wang, Tao Liu, Zhilin Pei, and Hengjie Li. FlashGS: Efficient 3D Gaussian Splatting for Large-scale and High-resolution Ren- dering.arXiv preprint arXiv:2408.07967, 2024. 3

    2024 arXiv

  7. [14]

    Learning an animatable detailed 3D face model from in-the- wild images.ACM Trans

    Yao Feng, Haiwen Feng, Michael J Black, and Timo Bolkart. Learning an animatable detailed 3D face model from in-the- wild images.ACM Trans. Graph., 40(4):1–13, 2021. 3

    2021

  8. [15]

    Plenox- els: Radiance Fields without Neural Networks

    Sara Fridovich-Keil, Alex Yu, Matthew Tancik, Qinhong Chen, Benjamin Recht, and Angjoo Kanazawa. Plenox- els: Radiance Fields without Neural Networks. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recog., pages 5501– 5510, 2022. 3

    2022

  9. [16]

    Accurate, Dense, and Robust Multiview Stereopsis.IEEE Trans

    Yasutaka Furukawa and Jean Ponce. Accurate, Dense, and Robust Multiview Stereopsis.IEEE Trans. Pattern Anal. Mach. Intell., 32(8):1362–1376, 2010. 2

    2010

  10. [17]

    Low-poly Mesh Generation for Building Models

    Xifeng Gao, Kui Wu, and Zherong Pan. Low-poly Mesh Generation for Building Models. InProc. ACM SIGGRAPH,

  11. [18]

    Heckbert

    Michael Garland and Paul S. Heckbert. Surface simplifica- tion using quadric error metrics. InProceedings of the 24th Annual Conference on Computer Graphics and Interactive Techniques, page 209–216, 1997. 3

    1997

  12. [19]

    Line-Based 3D Building Ab- straction and Polygonal Surface Reconstruction From Im- ages.IEEE Trans

    Jianwei Guo, Yanchao Liu, Xin Song, Haoyu Liu, Xiaopeng Zhang, and Zhanglin Cheng. Line-Based 3D Building Ab- straction and Polygonal Surface Reconstruction From Im- ages.IEEE Trans. Vis. Comput. Graph., pages 3283–3297,

  13. [21]

    2D Gaussian Splatting for Geometrically Ac- curate Radiance Fields

    Binbin Huang, Zehao Yu, Anpei Chen, Andreas Geiger, and Shenghua Gao. 2D Gaussian Splatting for Geometrically Ac- curate Radiance Fields. InProc. ACM SIGGRAPH, 2024. 7

    2024

  14. [22]

    A survey of simple geometric primitives de- tection methods for captured 3D data

    Adrien Kaiser, Jose Alonso Ybanez Zepeda, and Tamy Boubekeur. A survey of simple geometric primitives de- tection methods for captured 3D data. InComput. Graph. Forum, pages 167–196, 2019. 3

    2019

  15. [23]

    3D Gaussian Splatting for Real-Time Radiance Field Rendering.ACM Trans

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk ¨uhler, and George Drettakis. 3D Gaussian Splatting for Real-Time Radiance Field Rendering.ACM Trans. Graph., 42(4), 2023. 2, 3, 6, 7

    2023

  16. [24]

    A Hierarchical 3D Gaussian Representation for Real-Time Rendering of Very Large Datasets.ACM Transactions on Graphics, 43(4), 2024

    Bernhard Kerbl, Andreas Meuleman, Georgios Kopanas, Michael Wimmer, Alexandre Lanvin, and George Drettakis. A Hierarchical 3D Gaussian Representation for Real-Time Rendering of Very Large Datasets.ACM Transactions on Graphics, 43(4), 2024. 2, 3

    2024

  17. [25]

    Integrating meshes and 3d gaussians for indoor scene reconstruction with sam mask guidance.arXiv preprint arXiv:2407.16173, 2024

    Jiyeop Kim and Jongwoo Lim. Integrating meshes and 3d gaussians for indoor scene reconstruction with sam mask guidance.arXiv preprint arXiv:2407.16173, 2024. 3

    2024 arXiv

  18. [26]

    A Hybrid Multiview Stereo Algorithm for Modeling Urban Scenes.IEEE Trans

    Florent Lafarge, Renaud Keriven, Mathieu Br ´edif, and Hoang-Hiep Vu. A Hybrid Multiview Stereo Algorithm for Modeling Urban Scenes.IEEE Trans. Pattern Anal. Mach. Intell., 35(1):5–17, 2013. 3

    2013

  19. [27]

    Dreammesh4d: Video-to-4d generation with sparse-controlled gaussian- mesh hybrid representation

    Zhiqi Li, Yiming Chen, and Peidong Liu. Dreammesh4d: Video-to-4d generation with sparse-controlled gaussian- mesh hybrid representation. InAdvances in Neural Infor- mation Processing Systems (NeurIPS), 2024. 3

    2024

  20. [28]

    VastGaussian: Vast 3D Gaus- sians for Large Scene Reconstruction

    Jiaqi Lin, Zhihao Li, Xiao Tang, Jianzhuang Liu, Shiyong Liu, Jiayue Liu, Yangdi Lu, Xiaofei Wu, Songcen Xu, You- liang Yan, and Wenming Yang. VastGaussian: Vast 3D Gaus- sians for Large Scene Reconstruction. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recog., pages 5166–5175,...

    2024

  21. [29]

    CityGaussian: Real-time High-quality Large-Scale Scene Rendering with Gaussians

    Yang Liu, Chuanchen Luo, Lue Fan, Naiyan Wang, Jun- ran Peng, and Zhaoxiang Zhang. CityGaussian: Real-time High-quality Large-Scale Scene Rendering with Gaussians. InProc. Eur. Conf. Comput. Vis., pages 265–282, 2025. 3, 7

    2025

  22. [30]

    CityGaussianV2: Efficient and Geomet- rically Accurate Reconstruction for Large-Scale Scenes

    Yang Liu, Chuanchen Luo, Zhongkai Mao, Junran Peng, and Zhaoxiang Zhang. CityGaussianV2: Efficient and Geomet- rically Accurate Reconstruction for Large-Scale Scenes. In Proc. Int. Conf. Learn. Represent., 2025. 2, 3, 7

    2025

  23. [31]

    SMPL: A Skinned Multi- Person Linear Mode.ACM Trans

    Matthew Loper, Naureen Mahmood, Javier Romero, Gerard Pons-Moll, and Michael J Black. SMPL: A Skinned Multi- Person Linear Mode.ACM Trans. Graph., 39(5), 2015. 3

    2015

  24. [32]

    Scaffold-GS: Structured 3D Gaussians for View-Adaptive Rendering

    Tao Lu, Mulin Yu, Linning Xu, Yuanbo Xiangli, Limin Wang, Dahua Lin, and Bo Dai. Scaffold-GS: Structured 3D Gaussians for View-Adaptive Rendering. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recog., pages 20654– 20664, 2024. 3

    2024

  25. [33]

    Srinivasan, Matthew Tancik, Jonathan T

    Ben Mildenhall, Pratul P. Srinivasan, Matthew Tancik, Jonathan T. Barron, Ravi Ramamoorthi, and Ren Ng. NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis. InProc. Eur. Conf. Comput. Vis., 2020. 2, 3

    2020

  26. [34]

    Instant Neural Graphics Primitives with a Mul- tiresolution Hash Encoding.ACM Trans

    Thomas M ¨uller, Alex Evans, Christoph Schied, and Alexan- der Keller. Instant Neural Graphics Primitives with a Mul- tiresolution Hash Encoding.ACM Trans. Graph., 41(4),

  27. [35]

    ScaNeRF: Scalable Bundle-Adjusting Neural Ra- diance Fields for Large-Scale Scene Rendering.ACM Trans

    Thomas M ¨uller, Alex Evans, Christoph Schied, and Alexan- der Keller. ScaNeRF: Scalable Bundle-Adjusting Neural Ra- diance Fields for Large-Scale Scene Rendering.ACM Trans. Graph., 42(6), 2023. 3

    2023

  28. [36]

    Octree-gs: Towards consistent real-time rendering with lod-structured 3d gaussians.arXiv preprint arXiv:2403.17898, 2024

    Kerui Ren, Lihan Jiang, Tao Lu, Mulin Yu, Linning Xu, Zhangkai Ni, and Bo Dai. Octree-gs: Towards consistent real-time rendering with lod-structured 3d gaussians.arXiv preprint arXiv:2403.17898, 2024. 3, 7

    2024 arXiv

  29. [37]

    Sch ¨onberger and Jan-Michae Frahm

    Johannes L. Sch ¨onberger and Jan-Michae Frahm. Structure- from-Motion Revisited. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recog., pages 4104–4113, 2016. 2

    2016

  30. [38]

    SplattingAvatar: Realistic Real-Time Human Avatars with Mesh-Embedded Gaussian Splatting

    Zhijing Shao, Zhaolong Wang, Zhuang Li, Duotun Wang, Xiangru Lin, Yu Zhang, Mingming Fan, and Zeyu Wang. SplattingAvatar: Realistic Real-Time Human Avatars with Mesh-Embedded Gaussian Splatting. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recog., 2024. 3

    2024

  31. [39]

    Seitz, and Richard Szeliski

    Noah Snavely, Steven M. Seitz, and Richard Szeliski. Skele- tal graphs for efficient structure from motion. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recog., pages 1–8,

  32. [40]

    NOPE- SAC: Neural One-Plane RANSAC for Sparse-View Planar 3D Reconstruction.IEEE Trans

    Bin Tan, Nan Xue, Tianfu Wu, and Gui-Song Xia. NOPE- SAC: Neural One-Plane RANSAC for Sparse-View Planar 3D Reconstruction.IEEE Trans. Pattern Anal. Mach. Intell., 45(12):15233–15248, 2023. 3

    2023

  33. [41]

    Mildenhall, Pratul Srinivasan, Jonathan T

    Matthew Tancik, Vincent Casser, Xinchen Yan, Sabeek Prad- han, Ben P. Mildenhall, Pratul Srinivasan, Jonathan T. Bar- ron, and Henrik Kretzschmar. Block-NeRF: Scalable Large Scene Neural View Synthesis. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recog., pages 8238–8248, 2022. 3

    2022

  34. [42]

    Mega-NeRF: Scalable Construction of Large- Scale NeRFs for Virtual Fly- Throughs

    Haithem Turki, Deva Ramanan, and Mahadev Satya- narayanan. Mega-NeRF: Scalable Construction of Large- Scale NeRFs for Virtual Fly- Throughs. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recog., pages 12912–12921,

  35. [43]

    LOD Generation for Urban Scenes.ACM Trans

    Yannick Verdie, Florent Lafarge, and Pierre Alliez. LOD Generation for Urban Scenes.ACM Trans. Graph., 34(3),

  36. [44]

    PointNeRF++: A multi-scale, point-based Neural Radiance Field

    Sun Weiwei, Trulls Eduard, Tseng Yang-Che, Samban- dam Sneha, Gopal Sharma, Andrea Tagliasacchi, and Kwang Moo Yi. PointNeRF++: A multi-scale, point-based Neural Radiance Field. InEuropean Conference on Com- puter Vision, 2024. 3

    2024

  37. [45]

    TwinTex: Geometry-Aware Texture Generation for Abstracted 3D Ar- chitectural Models.ACM Trans

    Weidan Xiong, Hongqian Zhang, Botao Peng, Ziyu Hu, Yongli Wu, Jianwei Guo, and Hui Huang. TwinTex: Geometry-Aware Texture Generation for Abstracted 3D Ar- chitectural Models.ACM Trans. Graph., 42(6), 2023. 5

    2023

  38. [46]

    Point-nerf: Point- based neural radiance fields

    Qiangeng Xu, Zexiang Xu, Julien Philip, Sai Bi, Zhixin Shu, Kalyan Sunkavalli, and Ulrich Neumann. Point-nerf: Point- based neural radiance fields. In2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 5428–5438, 2022. 3

    2022

  39. [47]

    UrbanBIS: a Large-scale Benchmark for Fine-grained Urban Building Instance Segmentation

    Guoqing Yang, Fuyou Xue, Qi Zhang, Ke Xie, Chi-Wing Fu, and Hui Huang. UrbanBIS: a Large-scale Benchmark for Fine-grained Urban Building Instance Segmentation. In Proc. ACM SIGGRAPH, 2023. 7

    2023

  40. [48]

    MVSNet: Depth Inference for Unstructured Multi- view Stereo

    Yao Yao, Zixin Luo, Shiwei Li, Tian Fang, and Long Quan. MVSNet: Depth Inference for Unstructured Multi- view Stereo. InProc. Eur. Conf. Comput. Vis., pages 785– 801, 2018. 2

    2018

  41. [49]

    Finding good configurations of planar primitives in unorganized point clouds

    Mulin Yu and Florent Lafarge. Finding good configurations of planar primitives in unorganized point clouds. InProc. of the IEEE conference on Computer Vision and Pattern Recog- nition (CVPR), New Orleans, US, 2022. 5

    2022

  42. [50]

    Mip-Splatting: Alias-Free 3D Gaussian Splatting

    Zehao Yu, Anpei Chen, Binbin Huang, Torsten Sattler, and Andreas Geiger. Mip-Splatting: Alias-Free 3D Gaussian Splatting. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recog., pages 12912–12921, 2024. 3

    2024

  43. [51]

    Blending Surface Segmentation and Edit- ing for 3D Models.IEEE Trans

    Long Zhang, Jianwei Guo, Jun Xiao, Xiaopeng Zhang, and Dong-Ming Yan. Blending Surface Segmentation and Edit- ing for 3D Models.IEEE Trans. Vis. Comput. Graph., 28(8): 2879–2894, 2022. 3

    2022

  44. [52]

    On Scaling Up 3D Gaussian Splatting Training.arXiv preprint arXiv:2406.18533, 2024

    Hexu Zhao, Haoyang Weng, Daohan Lu, Ang Li, Jinyang Li, Aurojit Panda, and Saining Xie. On Scaling Up 3D Gaussian Splatting Training.arXiv preprint arXiv:2406.18533, 2024. 3 A. Building Point Cloud Completion Algo- rithm Here is the detailed algorithm of BPCC. Note that for th...

    2024

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.