Pith. sign in

REVIEW 3 major objections 6 minor 50 references

Proc-GS: Procedural Building Generation for City Assembly with 3D Gaussians

T0 review · 3 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Proc-GS couples procedural building code with 3D Gaussian Splatting, cutting synthetic model size about 4x at comparable rendering quality and enabling editable, scalable city assembly.

desk verdict Solid engineering result with a real dataset, but the 4x compression claim needs the base/variance split and scaling analysis before it stands as advertised. read the letter →

arxiv 2412.07660 v1 pith:KUM7XRIK submitted 2024-12-10 cs.CV

Yixuan Li , Xingjian Ran , Linning Xu , Tao Lu , Mulin Yu , Zhenzhi Wang , Yuanbo Xiangli , Dahua Lin
show 1 more author
Bo Dai
This is my paper · ORCID
classification cs.CV
keywords proceduralmodeling3DGaussianSplattingbuildingassetextractioncitygenerationnovelviewsynthesissparse-viewreconstructioncodeediting
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

Proc-GS claims that buildings, which normally appear in 3D Gaussian Splatting as one undifferentiated cloud of Gaussians, can instead be reconstructed as a small set of shared base assets driven by a procedural code. Under this scheme, every repeated window or corner is a single asset instantiated many times, and the optimizer updates all instances synchronously; small per-instance variance assets absorb real differences. The result, the paper argues, is rendering quality on par with unconstrained 3D-GS while reducing the Gaussian count by about a factor of four on synthetic scenes, plus strong robustness when only a handful of training views are available. Because the code string itself controls assembly, editing and recombination become as simple as editing text, and the assets can be laid out into whole city blocks. For real scenes, the code is obtained by annotating facades and projecting onto meshes, so the method still depends on human input even though the asset extraction itself is automatic.

What carries the argument

The load-bearing object is the procedural code string: a compact description such as "L1_C1, (L1_W1)*, L1_C2" that lists which base assets compose each level of a building. Each base asset is a set of 3D Gaussians defined in a local frame, with a bounding box, a pivot, and per-instantiation transformations; instantiating the code places copies of these Gaussians, and shared assets receive gradients from every copy. Per-instance variance assets, each also a small Gaussian set, let repeated copies differ without breaking sharing, and the Bbox Adaptive Clamp keeps Gaussians inside their asset boxes so extracted components can be recombined cleanly. This decomposition is what carries the compression, the sparse-view robustness, and the editability claims.

What would settle it

Take a real facade and train Proc-GS twice, once with a correct hand-annotated procedural code and once with the same code whose repeat boundaries are shifted by one window position; if rendering quality and asset extraction are nearly identical, the procedural constraint is not doing the work the paper claims, whereas a clear degradation would confirm that the code must be exactly right.

Watch

Extended reading notes

Core claim

The paper's central claim is that procedural code can be made the organizing structure of a 3D-GS reconstruction rather than an external add-on. Given a code string that says which base assets appear on each building level, Proc-GS initializes 3D Gaussians inside each asset's bounding box, instantiates the shared assets under rigid transformations, and optimizes the assembled building against rendering loss. The repeated assets are updated synchronously by gradients from all their appearances, and a separate variance asset per instance lets each copy deviate slightly in shape and color. With a bounding-box adaptive clamp to keep asset boundaries clean, this yields synthetic-scene results at 27.68 PSNR versus 27.54 for 3D-GS while using 291k Gaussians instead of 1,238k. The paper positions this as the first integration of procedural modeling with 3D-GS and as a route to editable, scalable city generation from both virtual and real captures.

Load-bearing premise

The method assumes a correct procedural code exists for every building, meaning someone must know, for each facade, which windows are the same asset and how they repeat; for real scenes the paper obtains this by manual facade annotation projected onto a mesh, so a wrong or approximate code would force visually different regions to share Gaussians and degrade both rendering and editing.

Editorial extensions

If this is right

  • On the synthetic MatrixBuilding benchmark, Proc-GS reaches a PSNR of 27.68 versus 27.54 for 3D-GS while cutting the Gaussian count from 1,238k to 291k, a roughly 4x model-size reduction at comparable quality.
  • Reducing training views from 469 to 24 or 47, Proc-GS stays far ahead of 3D-GS (19.70 versus 16.93 PSNR at 24 views), because every repeated asset is reinforced by all of its instances.
  • Editing a building becomes a code edit: swapping window assets, changing repeat counts, and assigning different variance assets produce new buildings without retraining.
  • Assembling assets from different source buildings into one city, guided by a rule-based layout generator, yields 3D-consistent city views that score lower on camera and depth error than the generation baselines compared in the paper.
  • Real-world buildings can be converted into the same reusable asset form through mesh extraction, facade annotation, and projection, with only a slight drop in rendering accuracy (27.19 versus 27.38 PSNR against 3D-GS).

Reading between the lines

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

  • The shared-asset prior is the real source of the compression: the method is betting that buildings are mostly repetitions, and the 4x model-size drop measures how much repetition exists in the dataset, not just the efficiency of the code representation.
  • A natural next step beyond the paper's scope is applying the same code-constrained Gaussian decomposition to other repeated structures, such as street furniture, industrial plants, or rows of trees, wherever a repeat pattern can be specified.
  • The variance-asset design suggests a testable prediction: on irregular real facades, most of the photorealism budget will be spent in the variance assets, so removing them should hurt real-scene PSNR more than synthetic-scene PSNR.
  • If procedural-code extraction ever becomes fully automatic, the pipeline converts ordinary drone photo sets into editable asset libraries, making the manual annotation step the main remaining bottleneck.
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 / 6 minor

Summary. The paper introduces Proc-GS, a framework that integrates procedural building code into 3D Gaussian Splatting (3D-GS). A building is decomposed into shared base assets plus per-instantiation variance assets, guided by a procedural code string. The authors introduce the MatrixBuilding dataset of 17 synthetic buildings with ground-truth procedural codes and multi-view images, and also apply the method to three real-world drone-captured scenes with manually annotated procedural codes. Experiments compare against vanilla 3D-GS for novel-view synthesis, report sparse-view results, and compare city-generation consistency (depth/camera error) against prior generative city models. The central claimed results are comparable rendering quality with roughly 4x fewer Gaussians on synthetic data, strong robustness to sparse views, and improved city-level geometric consistency.

Significance. If the claims hold, the paper makes a useful contribution by showing that procedural structure can be injected into 3D-GS optimization, enabling asset-level editing, reuse, and scalable city assembly. The MatrixBuilding dataset is a valuable resource for future work on structured building reconstruction. The sparse-view robustness is notable and well demonstrated. However, the headline compression claim is not cleanly established because the per-instantiation variance assets, which constitute the majority of the final Gaussian count, scale with the number of instantiations rather than with the number of shared assets. The city-generation comparison is also presented without a reproducible evaluation protocol. These gaps need to be addressed before the main claims can be fully accepted.

major comments (3)
  1. [Section 4.2, Table 1, and Section 3.3, Table 3] The factor-of-4 compression claim is not shown to be a consequence of procedural sharing. Table 3 (row 2) shows that the model with procedural code and clamp but without variance assets uses 87k Gaussians (PSNR 25.54), while the full model uses 291k (PSNR 27.68). Thus per-instantiation variance assets contribute roughly 204k of the 291k total, about 70% of the final model. Since Section 3.3 states that a separate variance asset is created for each instantiation of each base asset, the total Gaussian count should scale with the number of instantiations K, and the compression ratio relative to 3D-GS may collapse when per-instance variation is high—as Table 1's real-world result (500k vs. 384k, ~1.3x) already suggests. The paper does not report the base-asset/variance-asset split or the scaling of total Gaussians with K. The claim 'significantly reducing the model size by a factor of 4' should be qualified and supported by reporting this breakdown and a scaling analysis.
  2. [Section 4.4, Table 4, and Section 4.1 metrics] The city-generation comparison is not reproducible. The paper does not describe how the baseline methods (Persistent Nature, SceneDreamer, CityDreamer, GaussianCity) were evaluated: which generated city assets or scenes were used, what camera trajectories were employed, and whether the depth and camera error metrics were computed in the same coordinate space and with the same rendering resolution as for Proc-GS. Since these baselines are generative models while Proc-GS assembles assets from captured buildings, the two tasks are not directly comparable unless the protocol is carefully matched. Please provide the full evaluation protocol for all methods, or reframe the comparison as illustrative rather than a quantitative head-to-head.
  3. [Section 3.2 and Appendix B] The real-world pipeline is not automatic. The Introduction says procedural code may be obtained 'manually or using an off-the-shelf segmentation model,' but the actual method in Appendix B involves manually annotating 2D procedural code for each facade and projecting it onto the mesh; no segmentation model is used in the presented pipeline. The Limitations section does acknowledge this, but the main text should state explicitly that real-world code extraction requires human annotation, since this directly affects the scalability claim for real-world scenes. Please describe the level of human involvement accurately in the main text.
minor comments (6)
  1. [Table 1] The quantitative comparisons report only averages over scenes, with no standard deviations or per-scene breakdown. The real-world PSNR difference between 3D-GS and Proc-GS is 0.19 dB, which may be within run-to-run or scene-to-scene variation; adding error bars or per-scene results would strengthen the 'comparable quality' claim.
  2. [Figure 4] Figure 4 contains extraneous diagrams (labeled 'Sparse Voxel from SfM Points' and 'Neural Gaussian Prediction') that appear to be from a different method and do not correspond to the clamp operation described in the caption. Please replace these with a clean illustration of the Clamp Scale and Clamp Position operations.
  3. [Section 4.2, first paragraph] The sentence 'significantly reducing the model size by a factor of 4' refers only to the synthetic benchmark; the following sentence mentions the lower real-world compression. Consider stating both numbers in one place to avoid overgeneralization.
  4. [Section 3.4 and Appendix C] The building generator uses GPT-4o to convert raw procedural data into regular procedural code, but there is no evaluation of the correctness or robustness of this conversion. A small study (e.g., number of valid codes produced, human inspection results, or comparison to a deterministic rule-based converter) would help assess the reliability of the assembly stage.
  5. [Section 2.3] The related work on inverse procedural modeling is mentioned, but the paper does not clearly position Proc-GS relative to methods that infer procedural rules from images or 3D models. A short discussion of how the proposed code extraction (from known procedural codes or manual annotation) differs from inverse procedural modeling would improve the context.
  6. [Captions and typos] Figure 1 uses 'ProcGS' while the rest of the paper uses 'Proc-GS'; please standardize. Also, the caption of Figure 4 is incomplete and should describe both subfigures accurately.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the compression claim is empirical and externally benchmarked, not derived by construction.

full rationale

Proc-GS's derivation chain is self-contained against external baselines. The procedural code is an input, not a quantity fitted from the rendering loss; base and variance Gaussians are optimized under standard L1+SSIM supervision and evaluated with the same novel-view metrics as 3D-GS. The factor-of-4 Gaussian-count reduction in Table 1 is an empirical comparison, and the ablation showing variance assets contribute most of the 291k Gaussians (Table 3, rows 2 vs 4) is an honest measurement of where the parameters go; it may qualify the interpretation of the compression claim, but it does not make the claim true by construction. The sparse-view robustness is an experimentally observed property of shared-parameter optimization rather than a renamed fit. Self-citations such as MatrixCity [26] for the rendering protocol and [20, 48] for preprocessing tools are peripheral and not load-bearing. No equation equates an output with an input, and no fitted parameter is relabeled as a prediction. Therefore no significant circularity is present.

Assumptions & free parameters 4 free parameters · 4 assumptions · 1 invented entities

The central claim rests on the supplied procedural code and the assumption that buildings decompose into shared, similarity-transformed assets. The only invented component is the variance asset, which is a learned per-instance modulation and is supported only by internal ablations.

free parameters (4)
  • Initial Gaussian count N = 10000 for synthetic scenes
    Hand-chosen in Section 4.1; controls the density of the asset initialization.
  • Soft bounding box expansion margin = 20 cm
    Manually set in Section 4.1 to avoid over-clamping in Bbox Adaptive Clamp.
  • SSIM loss weight lambda_SSIM = 0.2
    Hand-chosen in Section 4.1 for the combined L1 + SSIM loss.
  • Facade thickness for real-world code projection = empirically set
    Appendix B states the z-range of bounding boxes is based on an empirically set facade thickness.
assumptions (4)
  • domain assumption Ground truth procedural code is available for each building.
    Sections 3.2 and 3.3 state this explicitly; without the code, the asset acquisition stage cannot start.
  • domain assumption A building can be represented as shared base assets transformed by similarity transforms plus per-instance variance.
    Section 3.3, Equation 5 assumes the procedural instantiation model exactly captures geometry; in real scenes this is approximate.
  • domain assumption The facade estimation and projection pipeline (2D-GS, planar primitive fitting, manual annotation) produces accurate 3D procedural code.
    Appendix B; if the projected code misaligns, Gaussians are shared across visually different regions.
  • standard math 3D-GS rasterization is differentiable and the L1+SSIM loss is sufficient to disentangle assets under the procedural constraint.
    Inherited from 3D-GS [22] and relied on in Section 3.3.
invented entities (1)
  • Variance assets
    purpose: Per-instantiation Gaussian sets that model appearance and geometry differences between repeated instances of a base asset.
    Only internal ablation (Table 3 row 4 vs row 3) supports their utility; no external or falsifiable evidence outside the paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Proc-GS: Procedural Building Generation for City Assembly with 3D Gaussians." pith.science (2026). https://pith.science/paper/KUM7XRIK

@misc{pith2026241207660,
  author       = {Pith},
  title        = {Pith review of: Proc-GS: Procedural Building Generation for City Assembly with 3D Gaussians},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KUM7XRIK}},
  note         = {Machine review of arXiv:2412.07660}
}
read the original abstract

Buildings are primary components of cities, often featuring repeated elements such as windows and doors. Traditional 3D building asset creation is labor-intensive and requires specialized skills to develop design rules. Recent generative models for building creation often overlook these patterns, leading to low visual fidelity and limited scalability. Drawing inspiration from procedural modeling techniques used in the gaming and visual effects industry, our method, Proc-GS, integrates procedural code into the 3D Gaussian Splatting (3D-GS) framework, leveraging their advantages in high-fidelity rendering and efficient asset management from both worlds. By manipulating procedural code, we can streamline this process and generate an infinite variety of buildings. This integration significantly reduces model size by utilizing shared foundational assets, enabling scalable generation with precise control over building assembly. We showcase the potential for expansive cityscape generation while maintaining high rendering fidelity and precise control on both real and synthetic cases.

Figures

Figures reproduced from arXiv: 2412.07660 by the authors.

Figure 1
Figure 1. Architectural structures in urban environments often exhibit repetitive patterns, such as the arrangement of windows and doors [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Example Building Procedural Code from City Sam￾ple [1]. Different levels are distinguished by colors, each repre￾sented by a string of characters indicating the instantiation of as￾sets: C E (external corner), P* (pillar), and W* (window). Base assets such as C E and W1 for Levels 1 and 2, shown on the right, are manually created by artists. Buildings are hierarchically decomposed into levels, where identical layers… view at source ↗
Figure 3
Figure 3. Overview of ProcGS. Our pipeline consists of two stages: (1) Asset Acquisition: We acquire the base assets in the training process of the 3D-GS. These assets are then assembled according to procedural code and add variance assets to create a complete building, which is used for novel view synthesis with Gaussian Splatting. (2) Asset Assembly: We use the building generator and city layout generator to assemble these … view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Bbox Adaptive Clamp. (1) Clamping Scale: if a Gaussian exceeds the soft bounding box, the scale of this Gaussian is reduced by half. (2) Clamping Position: repositions Gaussians that exceed the bounding box, realigning them within the bounding box. 4. Experiments 4.1. …
Figure 5
Figure 5. Figure 5: Qualitative results. The left section shows results from three real-world scenes, while the right section presents results from the MatrixBuilding dataset. Proc-GS achieves rendering quality comparable to 3D-GS. Green boxes in each image highlight pairs of instanti￾ati…
Figure 6
Figure 6. Figure 6: (1) Clamp effect on editing. Without clamp, the bound￾ary area of edited scene is intensively corrupted by artifacts, mak￾ing it impractical to create a new building with these assets; (2) Clamp effect on asset. We ablate effects of the clamp operation and demonstrate …
Figure 7
Figure 7. Figure 7: Qualitative comparisons of city generation results. Our Proc-GS framework learns base assets during the training process of 3D-GS using procedural codes, which are then manipulated to assemble these assets into a cohesive 3D city. Compared to other generation￾based met…
Figure 8
Figure 8. Figure 8: Dataset Overview. (a) Overview of the 17 buildings in our proposed MatrixBuilding dataset (b) Yellow cameras represent training views and purple cameras represent test views. The proportion of training views to test views is about 5:1. Building CHB CHD CHE CHF CHG CHH …
Figure 9
Figure 9. Figure 9: Extracting Procedural Code from Real-World Scenes. (a) We extract point clouds with good geometric structures from multi-view images using 2D-GS [20].Then we use the method [48] to automatically estimate the building facade. (b) For each facade, automatically render a …
Figure 10
Figure 10. Figure 10: An example of the prompt used to obtain regular procedural code. GPT-4o [21] takes the raw data, one or more examples as well as descriptions of procedural code as input and summarizes the regular procedural code as output. Proc-GS Dense Views Downsample 10x 3D-GS Dow…
Figure 11
Figure 11. Figure 11: Sparse View qualitative results. Proc-GS demonstrates significant robustness when reduces the number of training views, in contrast to 3D-GS [22], which exhibits a pronounced susceptibility to numerous artifacts under similar conditions. This difference is attributed …
Figure 12
Figure 12. Figure 12: Building Editing. (1) The upper part shows synthetic data results, where we arranged variance assets to spell our method’s name, emphasizing its high controllability. (2) The lower part presents three editing results from the real-world scene. 14 [PITH_FULL_IMAGE:fig…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 37 canonical work pages

  1. [1]

    2, 3, 4, 6, 11

    https://www.unrealengine.com/marketplace/product/city- sample. 2, 3, 4, 6, 11

    work page

  2. [2]

    https://www.sidefx.com/. 2

    work page

  3. [3]

    https://www.unrealengine.com/. 3, 11

    work page

  4. [4]

    Barron, Ben Mildenhall, Dor Verbin, Pratul P

    Jonathan T. Barron, Ben Mildenhall, Dor Verbin, Pratul P. Srinivasan, and Peter Hedman. Mip-nerf 360: Unbounded anti-aliased neural radiance fields. In CVPR, pages 5460–

    work page

  5. [5]

    Persistent nature: A generative model of un- bounded 3d worlds

    Lucy Chai, Richard Tucker, Zhengqi Li, Phillip Isola, and Noah Snavely. Persistent nature: A generative model of un- bounded 3d worlds. In CVPR, pages 20863–20874. IEEE,

    work page

  6. [6]

    Chan, Connor Z

    Eric R. Chan, Connor Z. Lin, Matthew A. Chan, Koki Nagano, Boxiao Pan, Shalini De Mello, Orazio Gallo, Leonidas J. Guibas, Jonathan Tremblay, Sameh Khamis, Tero Karras, and Gordon Wetzstein. Efficient geometry- aware 3d generative adversarial networks. In CVPR, pages 16102–16112. IEEE, 2022. 6

    work page 2022

  7. [7]

    Interactive procedural street modeling

    Guoning Chen, Gregory Esch, Peter Wonka, Pascal M ¨uller, and Eugene Zhang. Interactive procedural street modeling. ACM Trans. Graph., 27(3):103, 2008. 3

    work page 2008

  8. [8]

    Scene- dreamer: Unbounded 3d scene generation from 2d image collections

    Zhaoxi Chen, Guangcong Wang, and Ziwei Liu. Scene- dreamer: Unbounded 3d scene generation from 2d image collections. IEEE Trans. Pattern Anal. Mach. Intell., 45(12): 15562–15576, 2023. 8

    work page 2023

Show all 50 references
  1. [9]

    Procthor: Large-scale embodied ai using procedural generation

    Matt Deitke, Eli VanderBilt, Alvaro Herrasti, Luca Weihs, Kiana Ehsani, Jordi Salvador, Winson Han, Eric Kolve, Aniruddha Kembhavi, and Roozbeh Mottaghi. Procthor: Large-scale embodied ai using procedural generation. In NeurIPS, 2022. 3

    2022

  2. [10]

    Aliaga, and Bedrich Benes

    Ilke Demir, Daniel G. Aliaga, and Bedrich Benes. Procedu- ralization for editing 3d architectural models. In 3DV, pages 194–202. IEEE Computer Society, 2016. 3

    2016

  3. [11]

    Guibas, Noah Snavely, and Gordon Wetzstein

    Boyang Deng, Richard Tucker, Zhengqi Li, Leonidas J. Guibas, Noah Snavely, and Gordon Wetzstein. Streetscapes: Large-scale consistent street view generation using autore- gressive video diffusion. In SIGGRAPH (Conference Paper Track), page 27. ACM, 2024. 3

    2024

  4. [12]

    Citygen: Infinite and controllable 3d city layout generation

    Jie Deng, Wenhao Chai, Jianshu Guo, Qixuan Huang, Wen- hao Hu, Jenq-Neng Hwang, and Gaoang Wang. Citygen: Infinite and controllable 3d city layout generation. arXiv preprint arXiv:2312.01508, 2023. 1, 3

    2023 arXiv

  5. [13]

    Citycraft: A real crafter for 3d city generation.arXiv preprint arXiv:2406.04983, 2024

    Jie Deng, Wenhao Chai, Junsheng Huang, Zhonghan Zhao, Qixuan Huang, Mingyan Gao, Jianshu Guo, Shengyu Hao, Wenhao Hu, Jenq-Neng Hwang, Xi Li, and Gaoang Wang. Citycraft: A real crafter for 3d city generation.arXiv preprint arXiv:2406.04983, 2024. 3

    2024 arXiv

  6. [14]

    Gaussianeditor: Editing 3d gaussians delicately with text instructions

    Jiemin Fang, Junjie Wang, Xiaopeng Zhang, Lingxi Xie, and Qi Tian. Gaussianeditor: Editing 3d gaussians delicately with text instructions. arXiv preprint arXiv:2311.16037 ,

    arXiv

  7. [15]

    Kubric: A scalable dataset generator

    Klaus Greff, Francois Belletti, Lucas Beyer, Carl Doersch, Yilun Du, Daniel Duckworth, David J Fleet, Dan Gnanapra- gasam, Florian Golemo, Charles Herrmann, et al. Kubric: A scalable dataset generator. In CVPR, pages 3749–3761,

  8. [16]

    Inverse procedural modeling of branching structures by in- ferring l-systems

    Jianwei Guo, Haiyong Jiang, Bedrich Benes, Oliver Deussen, Xiaopeng Zhang, Dani Lischinski, and Hui Huang. Inverse procedural modeling of branching structures by in- ferring l-systems. ACM Trans. Graph., 39(5):155:1–155:13,

  9. [17]

    Instruct-nerf2nerf: Editing 3d scenes with instructions

    Ayaan Haque, Matthew Tancik, Alexei A Efros, Alek- sander Holynski, and Angjoo Kanazawa. Instruct-nerf2nerf: Editing 3d scenes with instructions. arXiv preprint arXiv:2303.12789, 2023. 2

    2023 arXiv

  10. [18]

    Liu He and Daniel G. Aliaga. COHO: context-sensitive city- scale hierarchical urban layout generation. arXiv preprint arXiv:2407.11294, 2024. 3

    2024 arXiv

  11. [19]

    Rushmeier, and Valentin Deschaintre

    Yiwei Hu, Paul Guerrero, Milos Hasan, Holly E. Rushmeier, and Valentin Deschaintre. Generating procedural materials from text or image prompts. In SIGGRAPH (Conference Pa- per Track), pages 4:1–4:11. ACM, 2023. 3

    2023

  12. [20]

    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. In SIGGRAPH (Conference Paper Track), page 32. ACM, 2024. 4, 11, 12

    2024

  13. [21]

    Gpt-4o system card

    Aaron Hurst, Adam Lerer, Adam P Goucher, Adam Perel- man, Aditya Ramesh, Aidan Clark, AJ Ostrow, Akila Weli- hinda, Alan Hayes, Alec Radford, et al. Gpt-4o system card. arXiv preprint arXiv:2410.21276, 2024. 5, 11, 13

    2024 arXiv

  14. [22]

    3d gaussian splatting for real-time radiance field rendering

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk ¨uhler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering. ACM Trans. Graph., 42(4):139:1– 139:14, 2023. 2, 3, 4, 5, 6, 11, 13

    2023

  15. [23]

    Procsy: Procedural synthetic dataset generation to- wards influence factor studies of semantic segmentation net- works

    Samin Khan, Buu Phan, Rick Salay, and Krzysztof Czar- necki. Procsy: Procedural synthetic dataset generation to- wards influence factor studies of semantic segmentation net- works. In CVPR Workshops, pages 88–96. Computer Vision Foundation / IEEE, 2019. 3

    2019

  16. [24]

    Michels, Bedrich Benes, S ¨oren Pirk, and Wojtek Palubicki

    Bosheng Li, Jonathan Klein, Dominik L. Michels, Bedrich Benes, S ¨oren Pirk, and Wojtek Palubicki. Rhizomorph: The coordinated function of shoots and roots. ACM Trans. Graph., 42(4):59:1–59:16, 2023. 3

    2023

  17. [25]

    Robus: A multimodal dataset for control- lable road networks and building layouts generation

    Tao Li, Ruihang Li, Huangnan Zheng, Shanding Ye, Shijian Li, and Zhijie Pan. Robus: A multimodal dataset for control- lable road networks and building layouts generation. arXiv preprint arXiv:2407.07835, 2024. 3

    2024 arXiv

  18. [26]

    Matrixcity: A large-scale city dataset for city-scale neural rendering and beyond

    Yixuan Li, Lihan Jiang, Linning Xu, Yuanbo Xiangli, Zhen- zhi Wang, Dahua Lin, and Bo Dai. Matrixcity: A large-scale city dataset for city-scale neural rendering and beyond. In ICCV, pages 3182–3192. IEEE, 2023. 6

    2023

  19. [27]

    Infinicity: Infinite-scale city synthesis

    Chieh Hubert Lin, Hsin-Ying Lee, Willi Menapace, Menglei Chai, Aliaksandr Siarohin, Ming-Hsuan Yang, and Sergey Tulyakov. Infinicity: Infinite-scale city synthesis. arXiv preprint arXiv:2301.09637, 2023. 1, 2

    2023 arXiv

  20. [28]

    Pro- cedural metamaterials: A unified procedural graph for meta- material design

    Liane Makatura, Bohan Wang, Yi-Lu Chen, Bolei Deng, Chris Wojtan, Bernd Bickel, and Wojciech Matusik. Pro- cedural metamaterials: A unified procedural graph for meta- material design. ACM Trans. Graph., 42(5):168:1–168:19,

  21. [29]

    Procedural 3d building reconstruction 9 using shape grammars and detectors

    Markus Mathias, Andelo Martinovic, Julien Weissenberg, and Luc Van Gool. Procedural 3d building reconstruction 9 using shape grammars and detectors. In 3DIMPVT, pages 304–311. IEEE Computer Society, 2011. 3

    2011

  22. [30]

    Example-based procedural modeling using graph grammars

    Paul Merrell. Example-based procedural modeling using graph grammars. ACM Trans. Graph. , 42(4):60:1–60:16,

  23. [31]

    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 syn- thesis. In ECCV (1), pages 405–421. Springer, 2020. 2, 3

    2020

  24. [32]

    Instant neural graphics primitives with a multires- olution hash encoding

    Thomas M ¨uller, Alex Evans, Christoph Schied, and Alexan- der Keller. Instant neural graphics primitives with a multires- olution hash encoding. ACM Trans. Graph. , 41(4):102:1– 102:15, 2022. 2

    2022

  25. [33]

    Gen Nishida, Adrien Bousseau, and Daniel G. Aliaga. Pro- cedural modeling of a building from a single image.Comput. Graph. Forum, 37(2):415–429, 2018. 3

    2018

  26. [34]

    Yoav I. H. Parish and Pascal M ¨uller. Procedural modeling of cities. In SIGGRAPH, pages 301–308. ACM, 2001. 3

    2001

  27. [35]

    Infinite photore- alistic worlds using procedural generation

    Alexander Raistrick, Lahav Lipson, Zeyu Ma, Lingjie Mei, Mingzhe Wang, Yiming Zuo, Karhan Kayan, Hongyu Wen, Beining Han, Yihan Wang, Alejandro Newell, Hei Law, Ankit Goyal, Kaiyu Yang, and Jia Deng. Infinite photore- alistic worlds using procedural generation. In CVPR, pages ...

    2023

  28. [36]

    Towards robust monocu- lar depth estimation: Mixing datasets for zero-shot cross- dataset transfer

    Ren ´e Ranftl, Katrin Lasinger, David Hafner, Konrad Schindler, and Vladlen Koltun. Towards robust monocu- lar depth estimation: Mixing datasets for zero-shot cross- dataset transfer. IEEE Trans. Pattern Anal. Mach. Intell., 44 (3):1623–1637, 2022. 6

    2022

  29. [37]

    Sch ¨onberger and Jan-Michael Frahm

    Johannes L. Sch ¨onberger and Jan-Michael Frahm. Structure- from-motion revisited. In CVPR, pages 4104–4113. IEEE Computer Society, 2016. 6

    2016

  30. [38]

    Large-scale terrain authoring through interac- tive erosion simulation

    Hugo Schott, Axel Paris, Lucie Fournier, Eric Gu ´erin, and Eric Galin. Large-scale terrain authoring through interac- tive erosion simulation. ACM Trans. Graph., 42(5):162:1– 162:15, 2023. 3

    2023

  31. [39]

    Ur- banworld: An urban world model for 3d city generation

    Yu Shang, Yuming Lin, Yu Zheng, Hangyu Fan, Jingtao Ding, Jie Feng, Jiansheng Chen, Li Tian, and Yong Li. Ur- banworld: An urban world model for 3d city generation. arXiv preprint arXiv:2407.11965, 2024. 1, 3

    2024 arXiv

  32. [40]

    Dreamgaussian: Generative gaussian splatting for effi- cient 3d content creation

    Jiaxiang Tang, Jiawei Ren, Hang Zhou, Ziwei Liu, and Gang Zeng. Dreamgaussian: Generative gaussian splatting for effi- cient 3d content creation. arXiv preprint arXiv:2309.16653,

    arXiv

  33. [41]

    Delicate textured mesh recovery from nerf via adaptive surface refinement

    Jiaxiang Tang, Hang Zhou, Xiaokang Chen, Tianshu Hu, Er- rui Ding, Jingdong Wang, and Gang Zeng. Delicate textured mesh recovery from nerf via adaptive surface refinement. arXiv preprint arXiv:2303.02091, 2023. 2

    2023 arXiv

  34. [42]

    Procedural modeling and physically based rendering for synthetic data generation in automotive appli- cations

    Apostolia Tsirikoglou, Joel Kronander, Magnus Wrenninge, and Jonas Unger. Procedural modeling and physically based rendering for synthetic data generation in automotive appli- cations. arXiv preprint arXiv:1710.06270, 2017. 3

    2017 arXiv

  35. [43]

    Bovik, Hamid R

    Zhou Wang, Alan C. Bovik, Hamid R. Sheikh, and Eero P. Simoncelli. Image quality assessment: from error visibility to structural similarity. IEEE Trans. Image Process., 13(4): 600–612, 2004. 5, 6

    2004

  36. [44]

    Gaussiancity: Generative gaussian splatting for unbounded 3d city generation

    Haozhe Xie, Zhaoxi Chen, Fangzhou Hong, and Ziwei Liu. Gaussiancity: Generative gaussian splatting for unbounded 3d city generation. arXiv preprint arXiv:2406.06526, 2024. 1, 3, 8

    2024 arXiv

  37. [45]

    Citydreamer: Compositional generative model of unbounded 3d cities

    Haozhe Xie, Zhaoxi Chen, Fangzhou Hong, and Ziwei Liu. Citydreamer: Compositional generative model of unbounded 3d cities. In CVPR, pages 9666–9675. IEEE, 2024. 1, 3, 6, 8

    2024

  38. [46]

    Physgaussian: Physics- integrated 3d gaussians for generative dynamics

    Tianyi Xie, Zeshun Zong, Yuxing Qiu, Xuan Li, Yutao Feng, Yin Yang, and Chenfanfu Jiang. Physgaussian: Physics- integrated 3d gaussians for generative dynamics. arXiv preprint arXiv:2311.12198, 2023. 2

    2023 arXiv

  39. [47]

    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. In CVPR, pages 5438–5448,

  40. [48]

    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. In CVPR, pages 6367–6376, 2022. 4, 11, 12

    2022

  41. [49]

    Efros, Eli Shecht- man, and Oliver Wang

    Richard Zhang, Phillip Isola, Alexei A. Efros, Eli Shecht- man, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. In CVPR, pages 586–

  42. [50]

    Cityx: Controllable procedural con- tent generation for unbounded 3d cities

    Shougao Zhang, Mengqi Zhou, Yuxi Wang, Chuanchen Luo, Rongyu Wang, Yiwei Li, Xucheng Yin, Zhaoxiang Zhang, and Junran Peng. Cityx: Controllable procedural con- tent generation for unbounded 3d cities. arXiv preprint arXiv:2407.17572, 2024. 3 10 Proc-GS: Procedural Building Gen...

    2024 arXiv

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.