Pith. sign in

REVIEW 3 major objections 6 minor 1 cited by

GaussianPainter: Painting Point Cloud into 3D Gaussians with Normal Guidance

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 ↗

Pith's one-line read This paper claims that normal-guided rotation resolves the non-uniqueness that makes feed-forward Gaussian prediction unstable, achieving the best compared novel-view results (PSNR 30.9) and cross-object appearance transfer.

desk verdict Normal-guided rotation is a genuine improvement, but the 30.9 vs 27.5 PSNR discrepancy between the main table and the robustness table undermines the headline result. read the letter →

arxiv 2412.17715 v1 pith:GJU63BBJ submitted 2024-12-23 cs.CV

Jingqiu Zhou , Lue Fan , Xuesong Chen , Linjiang Huang , Si Liu , Hongsheng Li This is my paper · ORCID
classification cs.CV
keywords 3DGaussiansplattingsurfacenormalpointcloudpaintingfeed-forwardgenerationnon-uniquenessnovelviewsynthesisappearancetransfertriplaneinjection
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

GaussianPainter aims to turn any point cloud into a renderable 3D Gaussian field — a set of points carrying rotation, scale, opacity, and color that can be rendered in real time — from a single reference image in one feed-forward pass. The paper's central argument is that feed-forward Gaussian prediction is unstable because the parameter space is non-unique: substantially different Gaussian fields can render nearly identical images, so a network cannot settle on consistent parameters. Its fix is to constrain each Gaussian's rotation with a predicted surface normal, reducing rotation to a 2-degree-of-freedom alignment and making the remaining parameters learnable. A multiscale triplane module injects the reference image's appearance into the point-cloud decoder. With this design the method reports the best compared novel-view synthesis results on a large object benchmark (PSNR 30.9, SSIM 0.945, LPIPS 0.134) and demonstrates transferring appearance between objects of different shapes.

What carries the argument

The load-bearing mechanism is normal-guided rotation: for each occupied point the network predicts a unit normal $n$, and the Gaussian's rotation matrix $R$ is defined as the rotation that aligns the world $z$-axis with $n$, with rotation axis $r = n \times z$ and angle $\theta = \angle(z, n)$, computed via the standard axis-angle rotation formula. This collapses the Gaussian rotation to two degrees of freedom and leaves the network to predict only scales, opacity, and spherical-harmonic colors. Normals are trained through an Isotropic Normal Rendering module that renders them into a 2D normal map and supervises with L1 and SSIM losses, avoiding the need for direct 3D normal labels. Appearance is injected separately by a multiscale triplane — three axis-aligned feature planes indexed by 3D position — built from self-supervised visual Transformer features, cross-attended with occupancy features, and added into the point-cloud decoder at several scales.

What would settle it

Train a second feed-forward model with identical capacity and data but unconstrained rotations and a stabilized training scheme; if it matches or exceeds PSNR 30.9 on the benchmark used in the main comparison, then the claimed non-uniqueness bottleneck is not the limiting factor and normal guidance is not necessary.

Watch

Extended reading notes

Core claim

The paper establishes that the rotation parameters, rather than opacity or spherical-harmonic colors, are the main source of non-uniqueness in Gaussian fields: a pilot study re-initializes rotations, scales, opacities, and SH coefficients while fixing locations, and reports instability scores showing rotations and scales diverge most across re-initializations. Because rotations and scales are entangled, GaussianPainter treats stable rotation as the key subproblem and proposes normal-guided rotation. Each point receives a surface normal estimated from occupancy features and rendered to a 2D normal map for supervision; the Gaussian's rotation is then defined by rotating the vertical axis onto that normal using the standard axis-angle formula. With rotation fixed, the network predicts only scales, opacity, and spherical-harmonic colors in a constrained space. The paper shows this outperforms isotropic Gaussian prediction and direct rotation prediction, and that the resulting model can transfer appearance across objects by learning part-level semantics.

Load-bearing premise

The method assumes that surface normals estimated from sparse occupancy features, supervised only through rendered 2D normal maps, are accurate enough to fix Gaussian rotations without losing the anisotropic expressiveness that makes 3D Gaussian splatting powerful.

Editorial extensions

If this is right

  • Feed-forward painting can replace per-object optimization for object-level Gaussian generation, turning any point cloud, whether scanned, generated, or noisy, into a renderable Gaussian field in one pass.
  • Anisotropic Gaussians with normal-guided rotation outperform isotropic Gaussians, so constraining rotation need not give up the expressiveness that makes 3D Gaussian splatting attractive.
  • Appearance transfer generalizes across objects with different shapes when part semantics are understood, enabling style transfer between unrelated 3D assets.
  • Multi-reference control and varying visual-feature layers allow style interpolation and diverse painting outcomes at inference time without retraining.
  • The method degrades gracefully on noisy point clouds, with rendered quality falling only modestly as the fraction of perturbed points rises to 90 percent.

Reading between the lines

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

  • A testable extension is to apply the same instability-score diagnostic to other feed-forward predictors, such as relighting or material parameters, to identify which parameter channels need geometric constraints.
  • Because the normal-guided rotation leaves one rotational degree of freedom unmodelled — roll around the normal — objects with strongly directional surface detail such as hair, fur, or brushed metal may require an additional per-point roll parameter.
  • The apparent semantic correspondence in cross-object transfer suggests the triplane injection is not merely copying 2D patches; probing with adversarial reference images could reveal whether part-level understanding is robust or an artefact of dataset bias.
  • If normals were supplied by geometry rather than learned, the method could factorize geometry and appearance, potentially painting point clouds into Gaussians without any reference image.
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 proposes GaussianPainter, a feed-forward method that converts a point cloud and a reference image into a 3D Gaussian field in a single forward pass. To address the non-uniqueness of Gaussian parameters, the method predicts a surface normal for each point and uses it to define the Gaussian rotation via Rodrigues' formula, leaving the network to predict scales, opacity, and color in a constrained space. A multiscale triplane module injects reference image appearance using DINOv2 features and cross-attention. The method is evaluated on OmniObject3D and Objaverse, reporting state-of-the-art novel view synthesis (PSNR 30.9 vs AGG 28.4 and TriplaneGaussian 25.8) and demonstrating qualitative cross-object appearance transfer.

Significance. If the reported numbers are reliable, the paper makes a useful contribution by showing that a feed-forward Gaussian predictor can be stabilized through normal-guidance, avoiding per-scene optimization while retaining anisotropic Gaussians. The upper-bound study (Table 5) is a thoughtful check on the expressiveness of the constrained rotation representation, and the code release supports reproducibility. However, the numerical self-consistency concerns detailed below must be resolved before the central SOTA claim can be accepted.

major comments (3)
  1. [Sec. 5.4, Table 4 vs Sec. 5.3, Table 1] Table 4 reports PSNR 27.5 for 0% noisy points, which is exactly the input of unperturbed GT point clouds, while Table 1 reports PSNR 30.9 for the full model on the OmniObject3D validation split. The center-crop protocol in Section 5.3 is described as applying to 'our experiments', so it should also apply to Table 4 unless otherwise stated. If both tables use the same split and protocol, the 3.4 dB gap is an internal inconsistency that invalidates at least one reported number; if they differ, that difference is not disclosed. Moreover, Table 5 reports 29.16 PSNR for per-scene optimized unconstrained rotation, which is lower than the feed-forward 30.9 in Table 1, suggesting further protocol discrepancies across tables. Please reconcile these numbers or explicitly state the evaluation protocol used in each table.
  2. [Sec. 5.4, Table 2 and surrounding text] The text says 'training the rotation prediction model from scratch is quite unstable and the model is hard to be well-converged. Thus, we utilize the pretrained weights of our normal-guided model as initialization and fine-tune the rotation prediction model.' Yet Table 2 lists both 'Rot. (from scratch)' and 'Rot. (fine-tuned)†'. The reader cannot tell whether a from-scratch model was actually trained and under what protocol, or whether the from-scratch row is a separate experiment. If the from-scratch result is included as evidence for the instability claim, its training setup (init, schedule, epochs) must be described. Please clarify the relationship between the text and the two rows.
  3. [Sec. 5.5, Table 5] The row labeled 'Normal-guided rotation (ours)' in Table 5 is obtained by optimizing normals jointly with the other Gaussian parameters for each scene, not by using the feed-forward normal predictor of GaussianPainter. As an upper-bound study this is a valid way to test the expressiveness of the rotation representation, but the phrase 'ours' is misleading because it does not reflect the end-to-end pipeline where the normal may be imperfect. Please rename the row and add a sentence stating that the feed-forward normal estimation quality is not evaluated by this experiment.
minor comments (6)
  1. [Sec. 3] The Instability Score is defined using 'standard variance' which seems to mean standard deviation or variance; the formula should use consistent terminology and specify whether the variance or its square root is used.
  2. [Sec. 4.1, Eq. (2)] The rotation axis r = n × z is degenerate when n is parallel to z (the vertical direction); the paper should specify how this case is handled numerically to avoid division by zero.
  3. [Sec. 5.3, compared methods] For the baseline comparisons, the paper states that AGG and TriplaneGaussian are reimplemented or run with GT point clouds substituted; please specify in Sec. 5.3 whether the Objaverse cross-object experiments also use the same substitution, and whether the numbers in Table 1 are obtained from the official code or the reimplementation.
  4. [Sec. 5.5, semantic understanding] The 'Semantic understanding in the texture injection' subsection relies entirely on qualitative inspection of a few examples and refers to 'OmniObject' instead of 'OmniObject3D'; the claim of semantic understanding would be stronger with a quantitative or at least more systematic evaluation.
  5. [Abstract and throughout] There are minor terminological and typographical issues: 'KeyWords' should be 'Keywords', 'multi-scale' is sometimes written 'multiscale', and 'OminiObject3D' appears instead of 'OmniObject3D'.
  6. [General] All quantitative comparisons are reported as single numbers without error bars or multiple-seed statistics; given the small gaps in some ablations (e.g., Table 3, multi-scale 30.6 vs full model 30.9), the authors should state whether these are single runs or averaged over seeds.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the normal-guided rotation constraint and the appearance injection are genuine architectural choices, and the headline results are benchmarked against external baselines.

full rationale

GaussianPainter's derivation chain is self-contained and externally benchmarked. The central design—predicting per-point normals via Isotropic Normal Rendering with supervision on 2D normal maps, then deriving Gaussian rotations from those normals via Rodrigues' formula (Sec. 4.1, Eq. 1-2)—is a genuine constraint on the Gaussian parameter space, not an input that already contains the output. The claimed motivation (normal guidance mitigates rotation non-uniqueness) is supported by the pilot study in Sec. 3, which measures an Instability Score on re-initialized optimized Gaussian fields, and by comparison against external baselines AGG and TriplaneGaussian in Table 1. Ablations in Table 2 compare against rotation prediction from scratch and from fine-tuned weights; the fine-tuned baseline is initialized from the normal-guided model, which weakens it as independent evidence, but the normal-guided model still outperforms it (30.9 vs 29.5) and the disclosed initialization does not make the comparison circular. Self-citations in Related Work (He et al. 2024; Huang et al. 2025) are illustrative, not load-bearing. No uniqueness theorem is imported from the authors' prior work, and no fitted parameter is renamed as a prediction. Note: the unresolved numerical gap between Table 1 (PSNR 30.9) and Table 4 (PSNR 27.5 at 0% noise) is a correctness/consistency concern, not a circularity, and does not affect the derivation-chain analysis.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The paper is a machine-learning method with no free parameters in the sense of a fitted formula; the listed free parameters are hand-chosen design choices that affect the central claim. The axioms are standard assumptions about 3DGS, normal estimation, DINOv2 features, and the causal role of non-uniqueness, each invoked to make the method work.

free parameters (3)
  • Occupancy grid resolution = 200x200x200
    The point cloud is voxelized into a 200^3 occupancy grid (Sec. 4.1), a hand-chosen resolution that affects geometric fidelity and memory.
  • Number of triplane scales = 4 scales (37, 74, 144, 288)
    Four triplane scales are used for appearance injection (Sec. 5.1), chosen without a reported systematic search.
  • Loss weighting between L1 and SSIM = not stated
    The rendering losses are L1 and SSIM (Sec. 4.1), but their relative weights are not given in the paper.
assumptions (5)
  • standard math 3D Gaussian splatting rendering and its differentiability (Kerbl et al. 2023) work as assumed.
    The entire pipeline rests on the standard 3DGS renderer; no derivation is provided in this paper.
  • domain assumption Surface normals can be estimated from occupancy features and supervised through 2D normal maps.
    Sec. 4.1 relies on this; the paper argues off-surface points can still get meaningful normals from rendering.
  • domain assumption Normal-guided rotation, which only rotates the z-axis to the normal, is a sufficient rotation parameterization for high-quality Gaussians.
    Table 5 tests this for optimized Gaussians (small drop vs unconstrained), but for the feed-forward model it remains an assumption.
  • domain assumption DINOv2 features provide appearance and semantic information transferable across objects.
    The triplane injection uses frozen DINOv2 features (Sec. 4.2); cross-object transfer depends on these features generalizing.
  • domain assumption The non-uniqueness of Gaussian fields, as measured by the pilot study's Instability Score, is the cause of feed-forward prediction difficulty.
    Sec. 3 defines IS and observes rotations are unstable; the causal link to training instability is inferred, not proven.

how reviews work

0 comments
Cite this review

Pith. "Pith review of GaussianPainter: Painting Point Cloud into 3D Gaussians with Normal Guidance." pith.science (2026). https://pith.science/paper/GJU63BBJ

@misc{pith2026241217715,
  author       = {Pith},
  title        = {Pith review of: GaussianPainter: Painting Point Cloud into 3D Gaussians with Normal Guidance},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GJU63BBJ}},
  note         = {Machine review of arXiv:2412.17715}
}
read the original abstract

In this paper, we present GaussianPainter, the first method to paint a point cloud into 3D Gaussians given a reference image. GaussianPainter introduces an innovative feed-forward approach to overcome the limitations of time-consuming test-time optimization in 3D Gaussian splatting. Our method addresses a critical challenge in the field: the non-uniqueness problem inherent in the large parameter space of 3D Gaussian splatting. This space, encompassing rotation, anisotropic scales, and spherical harmonic coefficients, introduces the challenge of rendering similar images from substantially different Gaussian fields. As a result, feed-forward networks face instability when attempting to directly predict high-quality Gaussian fields, struggling to converge on consistent parameters for a given output. To address this issue, we propose to estimate a surface normal for each point to determine its Gaussian rotation. This strategy enables the network to effectively predict the remaining Gaussian parameters in the constrained space. We further enhance our approach with an appearance injection module, incorporating reference image appearance into Gaussian fields via a multiscale triplane representation. Our method successfully balances efficiency and fidelity in 3D Gaussian generation, achieving high-quality, diverse, and robust 3D content creation from point clouds in a single forward pass.

Figures

Figures reproduced from arXiv: 2412.17715 by the authors.

Figure 1
Figure 1. Given any reference image, the GaussianPainter paints point clouds into 3D Gaussians in a feed-forward network. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The overview of GaussianPainter. Its first major component is [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Demonstration of different Gaussian fields and In [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: The illustration of converting the predicted normal [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: Qualitatively comparison for the task of cross [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Qualitative comparison between different rotation prediction strategies, using the OmniObject3D validation split [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: The results of GaussianPainter on generated point [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 8
Figure 8. Figure 8: Examples in OmniObject3D. The point clouds and reference images are from different boats. Note that the quality [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: Demonstration using features from different DINOv2 layers. The results show that different visual features could [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]
Figure 10
Figure 10. Figure 10: Demonstration of style control with two reference images. An intriguing property is the changing process abides by [PITH_FULL_IMAGE:figures/full_fig_p009_10.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. NanoGS: Training-Free Gaussian Splat Simplification

    cs.CV 2026-03 conditional novelty 6.0 of 10

    A training-free, CPU-based graph-merging method reduces 3D Gaussian Splat primitive counts by orders of magnitude while maintaining higher rendering fidelity than prior compaction methods.

Reference graph

Works this paper leans on

49 extracted references · 24 canonical work pages · cited by 1 Pith paper

  1. [1]

    Chen, Z.; Wang, F.; and Liu, H. 2023. Text-to-3d using gaussian splatting. arXiv preprint arXiv:2309.16585

    arXiv 2023

  2. [2]

    Chung, J.; Lee, S.; Nam, H.; Lee, J.; and Lee, K. M. 2023. Luciddreamer: Domain-free generation of 3d gaussian splatting scenes. arXiv preprint arXiv:2311.13384

    arXiv 2023

  3. [3]

    2019-12-8

    Dawson-Haggerty et al. 2019-12-8. trimesh. In https://trimesh.org/, 3.2.0

    work page 2019

  4. [4]

    Deitke, M.; Schwenk, D.; Salvador, J.; Weihs, L.; Michel, O.; VanderBilt, E.; Schmidt, L.; Ehsani, K.; Kembhavi, A.; and Farhadi, A. 2023. Objaverse: A universe of annotated 3d objects. In CVPR, 13142--13153

    work page 2023

  5. [5]

    Dosovitskiy, A.; Beyer, L.; Kolesnikov, A.; Weissenborn, D.; Zhai, X.; Unterthiner, T.; Dehghani, M.; Minderer, M.; Heigold, G.; Gelly, S.; et al. 2021. An image is worth 16x16 words: Transformers for image recognition at scale. In ICLR

    work page 2021

  6. [6]

    Gao, J.; Gu, C.; Lin, Y.; Zhu, H.; Cao, X.; Zhang, L.; and Yao, Y. 2023. Relightable 3d gaussian: Real-time point cloud relighting with brdf decomposition and ray tracing. arXiv preprint arXiv:2311.16043

    arXiv 2023

  7. [7]

    Gu \'e don, A.; and Lepetit, V. 2023. Sugar: Surface-aligned gaussian splatting for efficient 3d mesh reconstruction and high-quality mesh rendering. arXiv preprint arXiv:2311.12775

    arXiv 2023

  8. [8]

    A.; Holynski, A.; and Kanazawa, A

    Haque, A.; Tancik, M.; Efros, A. A.; Holynski, A.; and Kanazawa, A. 2023. Instruct-nerf2nerf: Editing 3d scenes with instructions. ICCV

    work page 2023

Show all 49 references
  1. [9]

    He, R.; Ma, K.; Huang, L.; Huang, S.; Gao, J.; Wei, X.; Dai, J.; Han, J.; and Liu, S. 2024. Freeedit: Mask-free reference-based image editing with multi-modal instruction. arXiv preprint arXiv:2409.18071

    2024 arXiv

  2. [10]

    Huang, L.; Fang, R.; Zhang, A.; Song, G.; Liu, S.; Liu, Y.; and Li, H. 2025. Fouriscale: A frequency perspective on training-free high-resolution image synthesis. In European Conference on Computer Vision, 196--212. Springer

    2025

  3. [11]

    Kerbl, B.; Kopanas, G.; Leimk \"u hler, T.; and Drettakis, G. 2023. 3D Gaussian Splatting for Real-Time Radiance Field Rendering. ACM Transactions on Graphics, 42(4)

    2023

  4. [12]

    Kondo, N.; Ikeda, Y.; Tagliasacchi, A.; Matsuo, Y.; Ochiai, Y.; and Gu, S. S. 2021. Vaxnerf: Revisiting the classic for voxel-accelerated neural radiance field. arXiv preprint arXiv:2111.13112

    2021 arXiv

  5. [13]

    Lassner, C.; and Zollhofer, M. 2021. Pulsar: Efficient sphere-based neural rendering. In CVPR, 1440--1449

    2021

  6. [14]

    Li, J.; Feng, Z.; She, Q.; Ding, H.; Wang, C.; and Lee, G. H. 2021. Mine: Towards continuous depth mpi with nerf for novel view synthesis. In ICCV, 12578--12588

    2021

  7. [15]

    Li, X.; Wang, H.; and Tseng, K.-K. 2023. Gaussiandiffusion: 3d gaussian splatting for denoising diffusion probabilistic models with structured noise. arXiv preprint arXiv:2311.11221

    2023 arXiv

  8. [16]

    Liang, Y.; Yang, X.; Lin, J.; Li, H.; Xu, X.; and Chen, Y. 2023. Luciddreamer: Towards high-fidelity text-to-3d generation via interval score matching. arXiv preprint arXiv:2311.11284

    2023 arXiv

  9. [17]

    Lin, C.-H.; Gao, J.; Tang, L.; Takikawa, T.; Zeng, X.; Huang, X.; Kreis, K.; Fidler, S.; Liu, M.-Y.; and Lin, T.-Y. 2023. Magic3d: High-resolution text-to-3d content creation. In CVPR, 300--309

    2023

  10. [18]

    Lionar, S.; Xu, X.; Lin, M.; and Lee, G. H. 2024. Nu-mcc: Multiview compressive coding with neighborhood decoder and repulsive udf. NeurIPS, 36

    2024

  11. [19]

    Liu, M.; Xu, C.; Jin, H.; Chen, L.; Varma T, M.; Xu, Z.; and Su, H. 2024. One-2-3-45: Any single image to 3d mesh in 45 seconds without per-shape optimization. NeurIPS, 36

    2024

  12. [20]

    V.; Tokmakov, P.; Zakharov, S.; and Vondrick, C

    Liu, R.; Wu, R.; Hoorick, B. V.; Tokmakov, P.; Zakharov, S.; and Vondrick, C. 2023 a . Zero-1-to-3: Zero-shot One Image to 3D Object. In ICCV

    2023

  13. [21]

    Liu, X.; Zhan, X.; Tang, J.; Shan, Y.; Zeng, G.; Lin, D.; Liu, X.; and Liu, Z. 2023 b . Humangaussian: Text-driven 3d human generation with gaussian splatting. arXiv preprint arXiv:2311.17061

    2023 arXiv

  14. [22]

    Luiten, J.; Kopanas, G.; Leibe, B.; and Ramanan, D. 2023. Dynamic 3d gaussians: Tracking by persistent dynamic view synthesis. arXiv preprint arXiv:2308.09713

    2023 arXiv

  15. [23]

    Melas-Kyriazi, L.; Laina, I.; Rupprecht, C.; and Vedaldi, A. 2023. Realfusion: 360deg reconstruction of any object from a single image. In CVPR, 8446--8455

    2023

  16. [24]

    P.; Tancik, M.; Barron, J

    Mildenhall, B.; Srinivasan, P. P.; Tancik, M.; Barron, J. T.; Ramamoorthi, R.; and Ng, R. 2021. Nerf: Representing scenes as neural radiance fields for view synthesis. Communications of the ACM, 65(1): 99--106

    2021

  17. [25]

    Nichol, A.; Jun, H.; Dhariwal, P.; Mishkin, P.; and Chen, M. 2022. Point-e: A system for generating 3d point clouds from complex prompts. arXiv preprint arXiv:2212.08751

    2022 arXiv

  18. [26]

    Oquab, M.; Darcet, T.; Moutakanni, T.; Vo, H.; Szafraniec, M.; Khalidov, V.; Fernandez, P.; Haziza, D.; Massa, F.; El-Nouby, A.; et al. 2023. Dinov2: Learning robust visual features without supervision. arXiv preprint arXiv:2304.07193

    2023 arXiv

  19. [27]

    J.; Florence, P.; Straub, J.; Newcombe, R.; and Lovegrove, S

    Park, J. J.; Florence, P.; Straub, J.; Newcombe, R.; and Lovegrove, S. 2019. Deepsdf: Learning continuous signed distance functions for shape representation. In CVPR, 165--174

    2019

  20. [28]

    T.; and Mildenhall, B

    Poole, B.; Jain, A.; Barron, J. T.; and Mildenhall, B. 2022. Dreamfusion: Text-to-3d using 2d diffusion. arXiv preprint arXiv:2209.14988

    2022 arXiv

  21. [29]

    Qian, Z.; Wang, S.; Mihajlovic, M.; Geiger, A.; and Tang, S. 2023. 3dgs-avatar: Animatable avatars via deformable 3d gaussian splatting. arXiv preprint arXiv:2312.09228

    2023 arXiv

  22. [30]

    Rombach, R.; Blattmann, A.; Lorenz, D.; Esser, P.; and Ommer, B. 2022. High-resolution image synthesis with latent diffusion models. In CVPR, 10684--10695

    2022

  23. [31]

    Ronneberger, O.; Fischer, P.; and Brox, T. 2015. U-net: Convolutional networks for biomedical image segmentation. In MICCAI, 234--241. Springer

    2015

  24. [32]

    Shih, M.-L.; Su, S.-Y.; Kopf, J.; and Huang, J.-B. 2020. 3d photography using context-aware layered depth inpainting. In CVPR, 8028--8038

    2020

  25. [33]

    Sun, Y.; Wang, Y.; Liu, Z.; Siegel, J.; and Sarma, S. 2020. Pointgrow: Autoregressively learned point cloud generation with self-attention. In WACV, 61--70

    2020

  26. [34]

    Tang, J.; Ren, J.; Zhou, H.; Liu, Z.; and Zeng, G. 2024. Dreamgaussian: Generative gaussian splatting for efficient 3d content creation. ICLR

    2024

  27. [35]

    A.; and Shakhnarovich, G

    Wang, H.; Du, X.; Li, J.; Yeh, R. A.; and Shakhnarovich, G. 2023. Score jacobian chaining: Lifting pretrained 2d diffusion models for 3d generation. In CVPR, 12619--12629

    2023

  28. [36]

    Wang, Z.; Lu, C.; Wang, Y.; Bao, F.; Li, C.; Su, H.; and Zhu, J. 2024. Prolificdreamer: High-fidelity and diverse text-to-3d generation with variational score distillation. NeurIPS, 36

    2024

  29. [37]

    Wu, T.; Zhang, J.; Fu, X.; Wang, Y.; Ren, J.; Pan, L.; Wu, W.; Yang, L.; Wang, J.; Qian, C.; et al. 2023. Omniobject3d: Large-vocabulary 3d object dataset for realistic perception, reconstruction and generation. In CVPR, 803--814

    2023

  30. [38]

    Xu, D.; Jiang, Y.; Wang, P.; Fan, Z.; Shi, H.; and Wang, Z. 2022 a . Sinnerf: Training neural radiance fields on complex scenes from a single image. In ECCV, 736--753. Springer

    2022

  31. [39]

    Xu, D.; Yuan, Y.; Mardani, M.; Liu, S.; Song, J.; Wang, Z.; and Vahdat, A. 2024. Agg: Amortized generative 3d gaussians for single image to 3d. arXiv preprint arXiv:2401.04099

    2024 arXiv

  32. [40]

    Xu, Q.; Xu, Z.; Philip, J.; Bi, S.; Shu, Z.; Sunkavalli, K.; and Neumann, U. 2022 b . Point-nerf: Point-based neural radiance fields. In CVPR, 5438--5448

    2022

  33. [41]

    Yan, Y.; Lin, H.; Zhou, C.; Wang, W.; Sun, H.; Zhan, K.; Lang, X.; Zhou, X.; and Peng, S. 2024. Street gaussians for modeling dynamic urban scenes. arXiv preprint arXiv:2401.01339

    2024 arXiv

  34. [42]

    Yang, G.; Huang, X.; Hao, Z.; Liu, M.-Y.; Belongie, S.; and Hariharan, B. 2019. Pointflow: 3d point cloud generation with continuous normalizing flows. In ICCV, 4541--4550

    2019

  35. [43]

    Yifan, W.; Serena, F.; Wu, S.; \"O ztireli, C.; and Sorkine-Hornung, O. 2019. Differentiable surface splatting for point-based geometry processing. ACM Transactions on Graphics (TOG), 38(6): 1--14

    2019

  36. [44]

    Yugay, V.; Li, Y.; Gevers, T.; and Oswald, M. R. 2023. Gaussian-slam: Photo-realistic dense slam with gaussian splatting. arXiv preprint arXiv:2312.10070

    2023 arXiv

  37. [45]

    Zhang, Y.; Wei, Y.; Jiang, D.; Zhang, X.; Zuo, W.; and Tian, Q. 2023. Controlvideo: Training-free controllable text-to-video generation. arXiv preprint arXiv:2305.13077

    2023 arXiv

  38. [46]

    Zhou, X.; Lin, Z.; Shan, X.; Wang, Y.; Sun, D.; and Yang, M.-H. 2023. Drivinggaussian: Composite gaussian splatting for surrounding dynamic autonomous driving scenes. arXiv preprint arXiv:2312.07920

    2023 arXiv

  39. [47]

    Zou, Z.-X.; Yu, Z.; Guo, Y.-C.; Li, Y.; Liang, D.; Cao, Y.-P.; and Zhang, S.-H. 2023. Triplane meets gaussian splatting: Fast and generalizable single-view 3d reconstruction with transformers. arXiv preprint arXiv:2312.09147

    2023 arXiv

  40. [48]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...

  41. [49]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

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