Toy-GS: Assembling Local Gaussians for Precisely Rendering Large-Scale Free Camera Trajectories

Junyan Su; Qi Liu; Xiaohan Zhang; Yukui Qiu; Zhenyu Sun

REVIEW 4 major objections 7 minor 38 references

Toy-GS: Assembling Local Gaussians for Precisely Rendering Large-Scale Free Camera Trajectories

T0 review · 4 major / 7 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 Toy-GS claims that splitting large free-trajectory scenes into pose-aligned regions and training a local 3D Gaussian Splatting model per region—then fusing local and global models at render time—raises rendering quality while cutting GPU…

desk verdict Toy-GS is a believable incremental step for 3DGS on free trajectories, but the unvalidated 1/N camera-selection threshold and the missing code/error bars keep me from fully trusting the numbers. read the letter →

arxiv 2412.10078 v1 pith:OKGUCLO2 submitted 2024-12-13 cs.CV

Xiaohan Zhang , Zhenyu Sun , Yukui Qiu , Junyan Su , Qi Liu This is my paper · ORCID

classification cs.CV

keywords 3DGaussianSplattingnovelviewsynthesislarge-scalescenesfreecameratrajectoriesadaptivespatialpartitioninglocal-globalrenderingmulti-viewconstraintsGPUmemoryreduction

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

Large-scale free camera trajectories—long, irregular paths with dense foreground and sparse background views—break whole-scene 3D Gaussian Splatting, both in quality and GPU memory. Toy-GS argues the cure is to divide the cameras and point cloud into regions by camera pose, train a separate Gaussian model per region, and then fuse local models at render time instead of merging them into one global model. On top of this, PatchMatch-style multi-view constraints reposition Gaussians to correct depth and geometry, and position-aware point adaptive control (PPAC) assigns larger Gaussians to distant areas. The paper reports that this combination raises PSNR by 1.35 dB on the Free dataset, 0.4 dB on Tanks and Temples advanced, and 1.19 dB on the new SCUTic dataset versus vanilla 3DGS, while cutting GPU memory by about 7 G. If correct, this makes high-quality rendering of long, messy camera paths feasible on a single GPU.

What carries the argument

The load-bearing structure is the adaptive spatial partitioning plus local-global rendering. K-means on camera poses yields regions $\{A_n\}_{n=1}^N$; a camera is kept for region $A_i$ iff the fraction of points it sees that lie in $A_i$ exceeds $1/N$ (Eq. 2), which is meant to be robust to varying point-cloud density because both numerator and denominator move together. Within each region, a separate 3D Gaussian Splatting model is trained with two additions: PatchMatch homography warping (Eqs. 5–6) between reference and source views to refine depth and normals and relocate Gaussians onto correct geometry, and PPAC, which scales Gaussian size by $\gamma(\mu)=\|\mu\|_2/r-1$ for $\|\mu\|_2 \geq 2r$ so distant regions are covered by larger splats. At render time, a new viewpoint first locates its region by k-means, then passes the distance test $\|C_{\text{new}}-O\|<\gamma$ and the visibility test $M'_{A_i}/M'_{c_{\text{new}}}>1/N$; if both hold, the local Gaussian renders it, otherwise the global Gaussian does. This fusion is what prevents the holes that arise when each region's Gaussian is simply merged and pruned as in VastGaussian.

What would settle it

On a free-trajectory scene, re-run training with the selection threshold in Eq. (2) varied (e.g., $1/(2N)$ and $2/N$) while keeping everything else fixed; if a non-$1/N$ threshold raises PSNR or removes holes in regions with sparse background coverage, the ratio test's critical assumption fails.

Watch

Extended reading notes

Core claim

On free camera trajectories, the spatial coverage of views is strongly inhomogeneous: foreground objects are captured densely and up close, while backgrounds are seen sparsely from far away, and parts of the scene may be invisible from every camera. Toy-GS's central claim is that this regime requires splitting the scene, not to save memory alone, but to let each local Gaussian model concentrate on the texture detail it can actually see, and that the split must follow the camera poses rather than axis-aligned boxes. It therefore clusters cameras with k-means on their poses, assigns each point in the sparse SfM cloud to a region, prunes and expands cameras per region using a visibility ratio test, trains a 3DGS per region in parallel, and then decides per new viewpoint, by distance and visibility constraints, whether to render with that region's local Gaussian or with a global Gaussian. The multi-view constraint (PatchMatch on rendered depth and normal maps) and PPAC (scale factor $\gamma(\mu)=\|\mu\|_2/r-1$ for points beyond radius $r$) are added to improve geometry and distant rendering. The reported result is consistent gains in PSNR/SSIM/LPIPS over 3DGS and over VastGaussian on two public datasets and the authors' SCUTic dataset, with memory usage that falls as region count rises.

Load-bearing premise

The camera-selection test assumes that the share of points a camera sees inside a region reliably tells whether that camera should train that region; on free trajectories where foreground is dense and background sparse, this threshold can drop informative cameras and leave holes.

Editorial extensions

If this is right

On the Free dataset, Toy-GS with three regions improves PSNR by 1.35 dB over 3DGS and by 1.37 dB over VastGaussian, with better SSIM and LPIPS.
On the Tanks and Temples advanced split, the method gains 0.4 dB over 3DGS (21.27 vs 20.87) and 1.16 dB over VastGaussian, while using about 1 G less memory than 3DGS.
On the SCUTic dataset, Toy-GS with three areas beats 3DGS by 1.19 dB overall (23.76 vs 22.57), beats VastGaussian by 1.79 dB, and cuts total GPU memory from about 16.4 G to 9.6 G.
Ablations show that PatchMatch+PPAC alone add 0.6 dB, the local-global renderer adds about 1.8 dB over global-only rendering, and increasing region count from 1 to 7 raises PSNR by 1.14 dB while reducing memory by about 6 G.
Rendering quality improves monotonically as the number of regions grows from 1 to 7, without the degradation seen in VastGaussian's global merging.

Reading between the lines

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

Because the camera-selection threshold $1/N$ is fixed by region count, scenes with extreme foreground/background imbalance may silently exclude cameras whose small in-region share is still geometrically essential; an adaptive threshold or a view-coverage prior is a natural next step, though untested in the paper.
The local-global fusion rule is only demonstrated for up to seven regions; at finer granularity the distance test $\|C_{\text{new}}-O\|<\gamma$ likely needs re-tuning, since the max-distance threshold grows with region spread and may admit the wrong local model.
Since the base is 3DGS, the same divide-train-fuse recipe could be dropped into other point-based or splatting renderers, and the SCUTic dataset (2–3 times larger than Free) could serve as a stress test for trajectory-level generalization.
The reported memory figures are peak GPU usage on one RTX 3090; the claimed 7 G saving is an average over benchmarks, so actual savings vary by scene and region count.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Toy-GS is a believable incremental step for 3DGS on free trajectories, but the unvalidated 1/N camera-selection threshold and the missing code/error bars keep me from fully trusting the numbers.

Read this as a system paper: it takes known pieces (k-means pose partitioning from F2-NeRF, regional training from VastGaussian, PatchMatch and PPAC from prior work) and assembles them into a pipeline that beats 3DGS and VastGaussian on the Free dataset by about 1.35 dB PSNR and on their new SCUTic dataset while using less memory. The local-global rendering rule (Eq. 8–9) is genuinely new, and the SCUTic dataset itself is a useful resource with over 600 images per category.

Credit where it is due: the tables are internally consistent across three datasets (PSNR, SSIM, LPIPS all improve), the ablation trend makes sense, and Figure 7 backs the memory/quality tradeoff. For a system paper, that is a solid level of evidence.

Now the soft spots, roughly in order of seriousness.

First, the 1/N threshold in Eq. 2 is never ablated. This is the rule that decides whether a camera gets to train a region, and the paper's own intro says free trajectories have dense foreground and sparse background. A camera pointed at a dense foreground object will often see only a small piece of a background region, fail the MAi/Mc > 1/N test, and be excluded from the region that needs more views. The authors claim the ratio is stable across densities, but they never vary the threshold or compare against a different selection rule. The reported gains might hold, but this is a genuine unknown.

Second, no error bars or seeds. A 0.4 dB gain on Tanks and Temples is small enough that a couple of unlucky runs could change the conclusion.

Third, the PatchMatch integration is underspecified. We are told it optimizes depth maps and then moves Gaussians to fix incorrect depths, but the exact update is not given. This is fixable in a page.

Fourth, the ablation prose has small mismatches with the table (e.g., 0.42 vs 0.042 for SSIM). Copyediting, but sloppy.

Fifth, no code or data link, and CityGaussian—mentioned in Related Work—is not compared.

Overall, the central idea is sound and the contribution is usable. I would send it to a serious referee now, but I would expect the review to insist on artifacts and an ablation of the selection threshold. If you are in large-scale 3DGS, read it and cite it.

Referee Report

4 major / 7 minor

Summary. The paper presents Toy-GS, a 3D Gaussian Splatting system for large-scale free camera trajectories. It partitions cameras and the point cloud with k-means on camera poses, expands or prunes cameras by a visibility ratio (Eq. 2), trains per-region 3DGS models, and enhances them with a PatchMatch-based multi-view constraint and position-aware point adaptive control (PPAC). A local-global rendering rule selects local Gaussians for new viewpoints that pass distance and visibility tests and falls back to a global Gaussian otherwise. The authors report state-of-the-art PSNR, SSIM, and LPIPS on the Free dataset, the Tanks and Temples advanced set, and a new SCUTic dataset, with GPU memory savings of several GB relative to 3DGS.

Significance. If the results hold, Toy-GS is a practical step toward applying 3DGS to trajectories that are neither object-centric nor aerial: it reduces memory through parallel region training and shows quality gains from region-local optimization plus a fusion strategy. The empirical evidence is broad, covering three datasets, and Figure 7 supports the quality-versus-region-count trend. The paper's value is as a system paper, not as a theoretical contribution, and it would be publishable once the central mechanisms are specified and the statistical uncertainty is addressed. The claimed gains are not yet fully supported because the PatchMatch-to-Gaussian update is unspecified, the camera-selection threshold is unablated, and the memory accounting is ambiguous.

major comments (4)

[Section 3.1, Eq. (2)] The 1/N camera-selection threshold is load-bearing but unvalidated. The paper's introduction states that free trajectories have dense foreground views and sparse background views. Under Eq. (2), a camera that sees many foreground points (large Mc) and only a few points in a background region Ai (small MAi) is discarded from exactly the region whose training data are already sparse. No ablation varies this threshold; Table 4 varies the number of areas, PatchMatch/PPAC, and global versus local-global rendering, but never the selection rule. Since the headline gains (e.g., 1.19 dB on SCUTic) depend on the quality of each local model, the authors should report sensitivity to the threshold or replace the rule with one that cannot reject informative sparse-background cameras.
[Section 3.2] The paper claims PatchMatch moves the Gaussian ellipsoids to their correct positions, but it does not specify the matching cost, the number or choice of source views, or how the optimized depth maps are converted into Gaussian mean updates. The only implementation details given are an iteration interval and a patch size in the Appendix. Because the components labeled 2P are ablated jointly in Table 4, the individual contribution of the PatchMatch constraint cannot be isolated, and the method cannot be reproduced or independently verified as written.
[Tables 1, 2, and 4; Section 3.3] The GPU-memory accounting is unclear. In Tables 1 and 2 the 'Ours (3 Areas)' memory values equal the VastGaussian values to the megabyte, and in Table 4 the RG and RLG rows have identical memory for the same number of areas. Section 3.3 states that viewpoints failing the local criteria are rendered by the global Gaussian used with VastGaussian, so the authors need to state whether that global model is trained and retained, and if so why its memory is not reflected. A per-component memory breakdown is needed to support the memory-saving claim in the abstract and Figure 7.
[Experimental protocol] All tables report a single run with no error bars, seeds, or statistical tests, and the train/test split is not defined for the SCUTic and Tanks and Temples evaluations. With headline differences as small as 0.4 dB (Table 2) and ablation differences of 0.02 in SSIM (Table 4), readers cannot assess whether the differences are within run-to-run variation. Please provide multi-seed results or confidence intervals for the main comparisons.

minor comments (7)

[Section 3.2, Eqs. (7)-(8)] The symbol gamma is used both for the PPAC scale factor and for the local-rendering distance threshold; please rename one to avoid confusion.
[Section 3.1, Eq. (1)] Equation (1) writes zp = KTP without defining the homogeneous coordinate convention; please state explicitly that z is the depth scale and that P and p are homogeneous vectors.
[Section 3.3] 'VastGaussain' is a typo for VastGaussian.
[Ablation Studies, Table 4 discussion] The text 'improves by 1.8 dB, 0.42 dB, and 0.032 dB in PSNR, SSIM, and LPIPS' should read 0.042 for SSIM, and the values are not all in dB.
[Table 4] The checkmarks for the RG and RLG columns are not visually distinguishable in the rendered text; please use explicit column entries so the comparisons described in the prose can be verified.
[Abstract] The '1.19 dB' gain should be identified as the SCUTic overall result rather than appearing to be an average over all benchmarks.
[SCUTic dataset] Please provide a public link, release plan, or access mechanism for the SCUTic dataset, since the paper bills the dataset as a contribution.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Toy-GS's rendering-quality claims are empirical measurements against external benchmarks, not consequences of a self-referential derivation.

full rationale

I walked the claimed derivation chain - adaptive spatial partitioning (Sec. 3.1), PatchMatch multi-view constraints and PPAC (Sec. 3.2), and local-global rendering (Sec. 3.3) - and found no step in which a prediction is equivalent to an input by construction. The camera-selection rule of Eq. (2) and the render-time assignment criteria of Eqs. (8)-(9) do reuse the same k-means pose model and the same 1/N visibility threshold, but this is a consistent partitioning scheme rather than a case of fitting a parameter and then reporting the fit as a prediction; the paper never claims that the PSNR gains are derived from these equations. PatchMatch is cited to Barnes et al. and PPAC to Chen et al., neither a self-citation, and the main quantitative claims are PSNR/SSIM/LPIPS measured on the public Free and Tanks and Temples datasets as well as the authors' SCUTic dataset. Using a self-collected dataset is not circular unless the model is tuned to that test set, which is not evidenced and is offset by the public-benchmark results. The ablation in Table 4 varies areas, PatchMatch/PPAC, and rendering mode, providing independent component-level evidence. A possible robustness concern about the 1/N camera-selection threshold excluding sparse-background cameras is a correctness risk, not a circularity, and no quoted equation reduces a claimed result to its own inputs.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The method depends on conventional 3DGS machinery plus several hand-set thresholds (k, r, 1/N, PatchMatch schedule) and unstated assumptions about pseudo-depth quality and global Gaussian construction. No new physical or representational entities are posited; the SCUTic dataset is a data contribution, not an entity.

free parameters (4)

number of regions k in k-means partition = 3 for main results; 5 and 7 in ablations
Chosen by hand in Section 3.1; no selection criterion or sensitivity analysis is given, and main experiments fix k=3 even though the ablation shows k=7 improves quality.
PPAC scene-radius threshold r = 0.005
Set in the appendix as 'the value of r is set to 0.005', but Equation 7 describes r as the scene radius; it is a hand-tuned scale controlling when larger Gaussians are used for distant content.
camera visibility threshold 1/N = 1/N, e.g., 1/3
Equations 2 and 9 accept a camera for a region only if MAi/Mc is greater than 1/N; this threshold is a modeling choice with no independent justification or robustness study.
PatchMatch patch size and optimization schedule = patch size 20, propagation every 50 iterations, active from 1k to 6k of 30k iterations
Hyperparameters listed in the appendix without sensitivity analysis.

assumptions (3)

domain assumption COLMAP sparse point cloud provides a reliable geometric prior for both partitioning and Gaussian initialization
The adaptive partition and 3DGS initialization in Section 3.1 both assume the sparse point cloud is dense enough and accurate enough to represent per-region visibility.
domain assumption Gaussian-rendered depth and normal maps are accurate enough for PatchMatch to produce useful multi-view depth constraints
Section 3.2 uses rendered depth and normal maps as input to PatchMatch and moves Gaussians based on the comparison; no validation of these pseudo-depth maps is provided.
ad hoc to paper A global Gaussian assembled from local models can render viewpoints that cross multiple regions without holes
Section 3.3 relies on 'the global Gaussian followed by VastGaussian' for viewpoints that fail the distance and visibility constraints; the construction, training, and failure modes of this global model are not specified.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Toy-GS: Assembling Local Gaussians for Precisely Rendering Large-Scale Free Camera Trajectories." pith.science (2026). https://pith.science/paper/OKGUCLO2

@misc{pith2026241210078,
  author       = {Pith},
  title        = {Pith review of: Toy-GS: Assembling Local Gaussians for Precisely Rendering Large-Scale Free Camera Trajectories},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OKGUCLO2}},
  note         = {Machine review of arXiv:2412.10078}
}

read the original abstract

Currently, 3D rendering for large-scale free camera trajectories, namely, arbitrary input camera trajectories, poses significant challenges: 1) The distribution and observation angles of the cameras are irregular, and various types of scenes are included in the free trajectories; 2) Processing the entire point cloud and all images at once for large-scale scenes requires a substantial amount of GPU memory. This paper presents a Toy-GS method for accurately rendering large-scale free camera trajectories. Specifically, we propose an adaptive spatial division approach for free trajectories to divide cameras and the sparse point cloud of the entire scene into various regions according to camera poses. Training each local Gaussian in parallel for each area enables us to concentrate on texture details and minimize GPU memory usage. Next, we use the multi-view constraint and position-aware point adaptive control (PPAC) to improve the rendering quality of texture details. In addition, our regional fusion approach combines local and global Gaussians to enhance rendering quality with an increasing number of divided areas. Extensive experiments have been carried out to confirm the effectiveness and efficiency of Toy-GS, leading to state-of-the-art results on two public large-scale datasets as well as our SCUTic dataset. Our proposal demonstrates an enhancement of 1.19 dB in PSNR and conserves 7 G of GPU memory when compared to various benchmarks.

Figures

Figures reproduced from arXiv: 2412.10078 by the authors.

**Figure 2.** The pipeline of Toy-GS. Firstly, we adaptively divide cameras and the point cloud into multiple areas based on the [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Adaptive camera selection strategy in an area. For [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (5 more)

**Figure 4.** Figure 4: Comparison of different rendering methods. Vast [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

**Figure 5.** Figure 5: Visual comparisons with recent methods on our SCUTic dataset. Our method provides better rendering effects for [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

**Figure 6.** Figure 6: Visual comparisons with recent methods on the [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

**Figure 7.** Figure 7: Left: The PSNR of the global rendering and local [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]

**Figure 8.** Figure 8: The camera trajectories in our SCUTic dataset. [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 19 canonical work pages

[1]

Aliev, K.-A.; Sevastopolsky, A.; Kolos, M.; Ulyanov, D.; and Lempitsky, V. 2020. Neural point-based graphics. In Computer Vision--ECCV 2020: 16th European Conference, Glasgow, UK, August 23--28, 2020, Proceedings, Part XXII 16, 696--712. Springer

work page 2020
[2]

Barnes, C.; Shechtman, E.; Finkelstein, A.; and Goldman, D. B. 2009. PatchMatch: A randomized correspondence algorithm for structural image editing. ACM Trans. Graph., 28(3): 24

2009
[3]

T.; Mildenhall, B.; Verbin, D.; Srinivasan, P

Barron, J. T.; Mildenhall, B.; Verbin, D.; Srinivasan, P. P.; and Hedman, P. 2022. Mip-nerf 360: Unbounded anti-aliased neural radiance fields. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 5470--5479

2022
[4]

Buehler, C.; Bosse, M.; McMillan, L.; Gortler, S.; and Cohen, M. 2001. Unstructured lumigraph rendering. In Proceedings of the 28th annual conference on Computer graphics and interactive techniques, 425--432

work page 2001
[5]

Chen, Y.; Gu, C.; Jiang, J.; Zhu, X.; and Zhang, L. 2023. Periodic vibration gaussian: Dynamic urban scene reconstruction and real-time rendering. arXiv preprint arXiv:2311.18561

arXiv 2023
[6]

Davis, A.; Levoy, M.; and Durand, F. 2012. Unstructured light fields. In Computer Graphics Forum, volume 31, 305--314. Wiley Online Library

work page 2012
[7]

E.; Taylor, C

Debevec, P. E.; Taylor, C. J.; and Malik, J. 2023. Modeling and rendering architecture from photographs: A hybrid geometry-and image-based approach. In Seminal Graphics Papers: Pushing the Boundaries, Volume 2, 465--474

work page 2023
[8]

Dhamo, H.; Tateno, K.; Laina, I.; Navab, N.; and Tombari, F. 2019. Peeking behind objects: Layered depth prediction from a single image. Pattern Recognition Letters, 125: 333--340

work page 2019

Show all 38 references

[9]

Flynn, J.; Broxton, M.; Debevec, P.; DuVall, M.; Fyffe, G.; Overbeck, R.; Snavely, N.; and Tucker, R. 2019. Deepview: View synthesis with learned gradient descent. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2367--2376

2019
[10]

J.; Grzeszczuk, R.; Szeliski, R.; and Cohen, M

Gortler, S. J.; Grzeszczuk, R.; Szeliski, R.; and Cohen, M. F. 2023. The lumigraph. In Seminal Graphics Papers: Pushing the Boundaries, Volume 2, 453--464

2023
[11]

He, T.; Collomosse, J.; Jin, H.; and Soatto, S. 2020. Deepvoxels++: Enhancing the fidelity of novel view synthesis from 3d voxel embeddings. In Proceedings of the Asian conference on computer vision

2020
[12]

Kerbl, B.; Kopanas, G.; Leimk \"u hler, T.; and Drettakis, G. 2023. 3D Gaussian Splatting for Real-Time Radiance Field Rendering. ACM Trans. Graph., 42(4): 139--1

2023
[13]

Knapitsch, A.; Park, J.; Zhou, Q.-Y.; and Koltun, V. 2017. Tanks and temples: Benchmarking large-scale scene reconstruction. ACM Transactions on Graphics (ToG), 36(4): 1--13

2017
[14]

Levin, A.; and Durand, F. 2010. Linear view synthesis using a dimensionality gap light field prior. In 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 1831--1838. IEEE

2010
[15]

Li, Z.; Xian, W.; Davis, A.; and Snavely, N. 2020. Crowdsampling the plenoptic function. In Computer Vision--ECCV 2020: 16th European Conference, Glasgow, UK, August 23--28, 2020, Proceedings, Part I 16, 178--196. Springer

2020
[16]

Lin, J.; Li, Z.; Tang, X.; Liu, J.; Liu, S.; Liu, J.; Lu, Y.; Wu, X.; Xu, S.; Yan, Y.; et al. 2024. Vastgaussian: Vast 3d gaussians for large scene reconstruction. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 5166--5175

2024
[17]

Liu, Y.; Guan, H.; Luo, C.; Fan, L.; Peng, J.; and Zhang, Z. 2024. Citygaussian: Real-time high-quality large-scale scene rendering with gaussians. arXiv preprint arXiv:2404.01133

2024 arXiv
[18]

Lombardi, S.; Simon, T.; Saragih, J.; Schwartz, G.; Lehrmann, A.; and Sheikh, Y. 2019. Neural volumes: Learning dynamic renderable volumes from images. arXiv preprint arXiv:1906.07751

2019 arXiv
[19]

Lombardi, S.; Simon, T.; Schwartz, G.; Zollhoefer, M.; Sheikh, Y.; and Saragih, J. 2021. Mixture of volumetric primitives for efficient neural rendering. ACM Transactions on Graphics (ToG), 40(4): 1--13

2021
[20]

P.; Ortiz-Cayon, R.; Kalantari, N

Mildenhall, B.; Srinivasan, P. P.; Ortiz-Cayon, R.; Kalantari, N. K.; Ramamoorthi, R.; Ng, R.; and Kar, A. 2019. Local light field fusion: Practical view synthesis with prescriptive sampling guidelines. ACM Transactions on Graphics (ToG), 38(4): 1--14

2019
[21]

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
[22]

M \"u ller, T.; Evans, A.; Schied, C.; and Keller, A. 2022. Instant neural graphics primitives with a multiresolution hash encoding. ACM transactions on graphics (TOG), 41(4): 1--15

2022
[23]

Shade, J.; Gortler, S.; He, L.-w.; and Szeliski, R. 1998. Layered depth images. In Proceedings of the 25th annual conference on Computer graphics and interactive techniques, 231--242

1998
[24]

Shih, M.-L.; Su, S.-Y.; Kopf, J.; and Huang, J.-B. 2020. 3d photography using context-aware layered depth inpainting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 8028--8038

2020
[25]

P.; Tucker, R.; Barron, J

Srinivasan, P. P.; Tucker, R.; Barron, J. T.; Ramamoorthi, R.; Ng, R.; and Snavely, N. 2019. Pushing the boundaries of view extrapolation with multiplane images. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 175--184

2019
[26]

Thies, J.; Zollh \"o fer, M.; and Nie ner, M. 2019. Deferred neural rendering: Image synthesis using neural textures. Acm Transactions on Graphics (TOG), 38(4): 1--12

2019
[27]

Tucker, R.; and Snavely, N. 2020. Single-view view synthesis with multiplane images. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 551--560

2020
[28]

Tulsiani, S.; Tucker, R.; and Snavely, N. 2018. Layer-structured 3d scene inference via view synthesis. In Proceedings of the European Conference on Computer Vision (ECCV), 302--317

2018
[29]

Turki, H.; Ramanan, D.; and Satyanarayanan, M. 2022. Mega-nerf: Scalable construction of large-scale nerfs for virtual fly-throughs. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 12922--12931

2022
[30]

Waechter, M.; Moehrle, N.; and Goesele, M. 2014. Let there be color! Large-scale texturing of 3D reconstructions. In Computer Vision--ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part V 13, 836--850. Springer

2014
[31]

Wang, P.; Liu, Y.; Chen, Z.; Liu, L.; Liu, Z.; Komura, T.; Theobalt, C.; and Wang, W. 2023. F2-nerf: Fast neural radiance field training with free camera trajectories. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 4150--4159

2023
[32]

N.; Azuma, D

Wood, D. N.; Azuma, D. I.; Aldinger, K.; Curless, B.; Duchamp, T.; Salesin, D. H.; and Stuetzle, W. 2023. Surface light fields for 3D photography. In Seminal Graphics Papers: Pushing the Boundaries, Volume 2, 487--496

2023
[33]

Xu, Q.; Xu, Z.; Philip, J.; Bi, S.; Shu, Z.; Sunkavalli, K.; and Neumann, U. 2022. Point-nerf: Point-based neural radiance fields. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 5438--5448

2022
[34]

Zhang, K.; Riegler, G.; Snavely, N.; and Koltun, V. 2020. Nerf++: Analyzing and improving neural radiance fields. arXiv preprint arXiv:2010.07492

2020 arXiv
[35]

A.; Shechtman, E.; and Wang, O

Zhang, R.; Isola, P.; Efros, A. A.; Shechtman, E.; and Wang, O. 2018. The unreasonable effectiveness of deep features as a perceptual metric. In Proceedings of the IEEE conference on computer vision and pattern recognition, 586--595

2018
[36]

Zhou, T.; Tucker, R.; Flynn, J.; Fyffe, G.; and Snavely, N. 2018. Stereo magnification: Learning view synthesis using multiplane images. arXiv preprint arXiv:1805.09817

2018 arXiv
[37]

, " * 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...
[38]

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.

[1] [1]

Aliev, K.-A.; Sevastopolsky, A.; Kolos, M.; Ulyanov, D.; and Lempitsky, V. 2020. Neural point-based graphics. In Computer Vision--ECCV 2020: 16th European Conference, Glasgow, UK, August 23--28, 2020, Proceedings, Part XXII 16, 696--712. Springer

work page 2020

[2] [2]

Barnes, C.; Shechtman, E.; Finkelstein, A.; and Goldman, D. B. 2009. PatchMatch: A randomized correspondence algorithm for structural image editing. ACM Trans. Graph., 28(3): 24

2009

[3] [3]

T.; Mildenhall, B.; Verbin, D.; Srinivasan, P

Barron, J. T.; Mildenhall, B.; Verbin, D.; Srinivasan, P. P.; and Hedman, P. 2022. Mip-nerf 360: Unbounded anti-aliased neural radiance fields. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 5470--5479

2022

[4] [4]

Buehler, C.; Bosse, M.; McMillan, L.; Gortler, S.; and Cohen, M. 2001. Unstructured lumigraph rendering. In Proceedings of the 28th annual conference on Computer graphics and interactive techniques, 425--432

work page 2001

[5] [5]

Chen, Y.; Gu, C.; Jiang, J.; Zhu, X.; and Zhang, L. 2023. Periodic vibration gaussian: Dynamic urban scene reconstruction and real-time rendering. arXiv preprint arXiv:2311.18561

arXiv 2023

[6] [6]

Davis, A.; Levoy, M.; and Durand, F. 2012. Unstructured light fields. In Computer Graphics Forum, volume 31, 305--314. Wiley Online Library

work page 2012

[7] [7]

E.; Taylor, C

Debevec, P. E.; Taylor, C. J.; and Malik, J. 2023. Modeling and rendering architecture from photographs: A hybrid geometry-and image-based approach. In Seminal Graphics Papers: Pushing the Boundaries, Volume 2, 465--474

work page 2023

[8] [8]

Dhamo, H.; Tateno, K.; Laina, I.; Navab, N.; and Tombari, F. 2019. Peeking behind objects: Layered depth prediction from a single image. Pattern Recognition Letters, 125: 333--340

work page 2019