Pith. sign in

REVIEW 5 major objections 6 minor 64 references

Monocular Dynamic Gaussian Splatting: Fast, Brittle, and Scene Complexity Rules

T0 review · 5 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 This benchmark of monocular dynamic Gaussian-splatting methods argues that the field's speed claims are real—20 to 200 frames per second—but that image quality is consistently beaten by a slower voxel-based method, TiNeuVox, and that…

desk verdict A genuinely useful reality-check benchmark for monocular dynamic Gaussian splatting, with one real caveat: the fairness of the cross-method comparison rests on an asserted but unshown reimplementation parity. read the letter →

arxiv 2412.04457 v2 pith:3WH7T4TC submitted 2024-12-05 cs.CV

Yiqing Liang , Mikhail Okunev , Mikaela Angelina Uy , Runfeng Li , Leonidas Guibas , James Tompkin , Adam W. Harley This is my paper · ORCID
classification cs.CV
keywords monoculardynamicviewsynthesis3DGaussiansplattingbenchmarkingmotionrepresentationadaptivedensitycontrolscenecomplexityspecularsurfacesimagequalitymetrics
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

This paper attempts to settle a crowded field: many monocular dynamic Gaussian splatting (GS) methods were published simultaneously, each claiming to be best, and the authors argue that cannot all be true. They integrate five representative GS methods—per-Gaussian polynomial/Fourier motion (EffGS), per-Gaussian polynomial/RBF motion (STG), shared MLP deformation field (DeformableGS), shared HexPlane field (4DGS), and direct 4D Gaussians (RTGS)—into one codebase and compare them on 50 real scenes across five datasets plus a purpose-built synthetic dataset. They find that all Gaussian methods render fast (20–200 FPS) but are consistently beaten in image quality by TiNeuVox, a slower voxel-based neural field, and that no Gaussian method achieves a stable rank order across real datasets. The paper argues that scene complexity—camera baseline, motion magnitude, pose accuracy, specular surfaces, and the fraction of static background—overwhelms the differences between motion representations, and that adaptive density control makes optimization brittle. The contribution is an apples-to-apples comparison and a set of findings that reframe what should be measured when progress is claimed.

What carries the argument

The central mechanism is a unified evaluation harness: five motion-representation families implemented in one codebase with shared hyperparameters, plus static 3DGS and the voxel-based TiNeuVox as baselines, run on a total of 50 scenes from five datasets and on a new instructive synthetic dataset. The synthetic dataset varies camera baseline B and object-motion distance D independently, isolating how each factor hurts reconstruction, while tracking-derived dynamic masks force separate measurement of quality on moving regions rather than letting static background dominate the metrics. This harness is what lets the paper attribute performance differences to motion representation rather than to dataset split or implementation details, and it is also what exposes the brittleness of adaptive density control.

What would settle it

Run the paper's five methods in their original released implementations with their original per-scene hyperparameters on the iPhone dataset and the narrow-baseline synthetic scenes, then compare masked LPIPS on dynamic regions; if a per-Gaussian method (EffGS or STG) or the 4D method RTGS beats DeformableGS and TiNeuVox, the paper's claims that field-based representations help and that scene complexity dominates would be contradicted.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that monocular dynamic Gaussian splatting is currently fast but brittle, and that scene difficulty rules. Averaged over five real-world datasets totaling 50 scenes, the non-Gaussian voxel method TiNeuVox reaches 24.54 PSNR while the best Gaussian method (DeformableGS) reaches 24.07, and TiNeuVox trains in comparable time and converges reliably; Gaussian methods render at 20–200 FPS versus TiNeuVox's 0.3 FPS. Among Gaussian methods, field-based motion representations (MLP or HexPlane) give better LPIPS than per-Gaussian low-order motion models, and the most expressive representation—direct 4D Gaussians—performs worst in both quality and efficiency. Yet across datasets the winning method changes, so no clear rank order emerges; on strictly-monocular iPhone data all Gaussian methods lag TiNeuVox by a wide margin. A controlled synthetic dataset shows reconstruction degrades as camera baseline shrinks or object motion grows, that static 3DGS can look competitive by reconstructing the background while inserting pseudo-dynamic Gaussians along a motion path, and that adaptive density control causes overfitting, varying efficiency, and occasional catastrophic optimization failures.

Load-bearing premise

The apples-to-apples comparison assumes that the integrated single-codebase reimplementations faithfully reproduce each original method, so the measured rank order reflects motion representation rather than implementation or hyperparameter differences.

Editorial extensions

If this is right

  • Published claims that a new motion representation is universally best should be treated as unestablished unless the method is tested across datasets like D-NeRF, Nerfies, HyperNeRF, NeRF-DS, and iPhone with masked dynamic-region metrics.
  • Field-based motion (shared MLP or HexPlane) is a safer default than per-Gaussian polynomial/Fourier motion or direct 4D Gaussians when scenes are monocular with narrow baselines.
  • A static 3DGS baseline can look competitive on full-image metrics even when it fails on moving objects, so masked metrics on dynamic regions are necessary for meaningful comparison.
  • Efficiency claims are not stable across scenes because adaptive density control produces very different Gaussian counts, so training and rendering times need per-scene reporting.
  • TiNeuVox's quality advantage suggests that hybrid voxel-field representations remain a viable alternative to Gaussian splatting where rendering speed is not the priority.

Reading between the lines

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

  • Editorial: The paper's measurements could be converted into a scalar scene-difficulty score combining baseline, motion magnitude, pose noise, and frequency content; if such a score predicted each method's error across the 50 scenes, benchmark findings would become transferable to new videos.
  • Editorial: If scene complexity dominates method choice, an adaptive model selector that estimates baseline and motion speed from a video and then picks the motion representation could outperform any fixed choice; the paper's released data could train such a selector.
  • Editorial: The paper's negative results for per-Gaussian and 4D methods should be read as "under uniform optimization," not "never works," because the shared-hyperparameter design is deliberate and a practitioner with per-scene tuning might obtain different rankings.
  • Editorial: The brittleness of adaptive density control points to densification scheduling and regularization, rather than new motion representations, as likely sources of near-term progress in monocular dynamic Gaussian splatting.
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

5 major / 6 minor

Summary. This manuscript benchmarks monocular dynamic view synthesis under Gaussian splatting. The authors collect 50 real-world scenes across D-NeRF, Nerfies, HyperNeRF, NeRF-DS, and iPhone, add a new synthetic dataset that varies camera baseline and object motion, and integrate five Gaussian methods (EffGS, STG, DeformableGS, 4DGS, RTGS) together with static 3DGS and TiNeuVox baselines into a shared codebase. They report three optimization runs per condition, standard deviations, and masked dynamic-region metrics. The paper's central thesis is that monocular dynamic Gaussian splatting is fast but brittle, that scene complexity and dataset factors overwhelm motion-representation differences, and that hybrid voxel/neural-field methods such as TiNeuVox remain competitive in quality. Eight findings summarize the evidence.

Significance. If the findings are correct, this is a useful and timely empirical contribution. Its strongest assets are the full-slate comparison on datasets where most original papers reported only subsets (Table 8), the controlled synthetic dataset with baseline/motion ablations, the use of three seeds and standard deviations, the masked static/dynamic metrics, and the public release of code and data. The paper is also appropriately candid about its limitations, including the lack of a method-independent scene-difficulty metric and the omission of concurrent works. The main risk to significance is reimplementation fidelity: the rank-order conclusions are only as trustworthy as the authors' integrated implementations, and that is not currently demonstrated.

major comments (5)
  1. [Sec. 4, first paragraph; Sec. 3.3; App. C.3; Table 2] The central assumption stated in Section 4, first paragraph, is that the integrated implementations 'approximately match the results of the original works', but the paper provides no parity table or quantitative evidence for this claim. This is especially consequential for STG: Section 3.3 and Appendix C.3 explain that STG was originally designed for multi-view input and had to be adapted for monocular input by copying the point cloud 10x along the time axis, and Table 2 reports a 4.56 dB gap between STG and DeformableGS (19.51 vs. 24.07). Because the qualitative conclusions are rank-order claims, even a 2-3 dB implementation gap on one method could change the story (compare TiNeuVox 24.54 and DeformableGS 24.07 in the same table). Please add a per-method, per-dataset parity check against the original papers' reported numbers on the scenes and splits they used, and justify or clearly separate the STG modifications and the STG/STG-decoder distinction.
  2. [Sec. 4.4; Table 2; Fig. 14] Section 4.4 states that collapsed optimization runs are excluded from the evaluation statistics, but the number, frequency, and location of these failures are never reported. This is load-bearing because Table 2 and Figure 5 are the empirical basis for both the quality rankings and the 'brittleness' claim. If some methods collapse more often, their reported means are optimistic and the comparison is not apples-to-apples. Please report failure counts per method and dataset, and provide a sensitivity analysis that either includes failed runs with a worst-case score (as already done for STG in Figure 14) or analyzes successful and failed runs separately.
  3. [Sec. 3.1; App. B.2; App. C.4; Table 9] The evaluation protocol for HyperNeRF is inconsistent between the main text and the appendix. Section 3.1 says original train/test splits are used, but Appendix B.2 reports that HyperNeRF has known bad camera poses, that the authors reran COLMAP with dynamic and specular regions masked, and that poses were improved in 7 of 17 scenes; Appendix C.4 says the Dynamic 3D Gaussian implementation uses the improved poses. Table 9 shows that pose correction improves some methods but degrades TiNeuVox, DeformableGS, and 4DGS. The main results in Table 2 and Figure 5 should state explicitly which pose set is used for HyperNeRF, and the rank-order analysis should be shown under both original and corrected poses so that cross-dataset findings do not mix protocols.
  4. [Table 2; Tables 3-7; Finding 2] Table 2 is the basis for Finding 2, but its aggregation rule is not specified. The appendix tables show that the five datasets have different sizes (D-NeRF 8 scenes, Nerfies 4, HyperNeRF 17, NeRF-DS 7, iPhone 14) and different winning methods across datasets. If Table 2 is an unweighted mean of dataset means, small datasets receive equal weight; if it is a scene-level mean, iPhone dominates. Since per-dataset winners differ (Figure 5), the aggregate rank order can be an artifact of the weighting choice. Please report the aggregation procedure and show whether Finding 2's ordering (field-based better than per-Gaussian, 4D worst) is stable under both conventions and across metrics.
  5. [Sec. 4.2; Table 1; Table 2] Finding 2's conclusion that 'going to 4D makes things worse' is based on a single representative of the 4D representation class, RTGS. Table 1 lists Rotor-Based 4DGS under the same '4D' type, and it is not evaluated. With only one 4D method, and one that the authors modified to fit the monocular setup, the claim about 4D representations generally is too strong. Please either include a second 4D method or restrict the conclusion to RTGS.
minor comments (6)
  1. [Secs. 2.2, 3.1, 4.4, 4.7] There are several typos: 'Liuten et al.' should be 'Luiten et al.' (Section 2.2), 'camaera' should be 'camera' (Section 3.1), 'substract' should be 'subtract' (Section 4.4), and 'outpeforms' should be 'outperforms' (Section 4.7).
  2. [Fig. 14; Figs. 30, 32, 34] The captions for the synthetic-dataset rankings describe 'solid bar' and 'textured bar', but the definitions are easy to confuse. Define them directly in the caption (e.g., all baselines vs. wide baselines) and apply the same convention consistently in the related appendix figures.
  3. [Table 2] The caption says 'averaged across all five datasets' but does not state whether this is the mean over dataset-level averages or over scenes. State the weighting explicitly and include standard deviations for the aggregate values, not only for per-dataset tables.
  4. [Sec. 3.3; Table 2] Section 3.3 introduces 'STG' and 'STG-decoder' but Table 2 reports both. Please clarify which version corresponds to the intended SpaceTimeGaussians method and why the decoder-removed variant is part of the headline comparison.
  5. [App. C.1; Sec. 3.3] Appendix C.1 lists differences from default 3DGS hyperparameters (batch size, warm-up, opacity reset, initialization). The phrase 'shared hyperparameters' in Section 3.3 should be qualified, since these differences are part of the comparison protocol and are not identical across all methods.
  6. [Sec. 6; Fig. 16] The limitations paragraph is honest about the lack of a method-independent scene-difficulty metric, but the claim that DyCheck's omega does not correlate is based on scatterplot inspection in Figure 16. Reporting a quantitative correlation coefficient would make the point precise.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is an empirical benchmark whose findings are measured from external datasets and third-party methods, not derived from its own fitted inputs or self-citations.

full rationale

This is an empirical benchmark paper. Its central claims—that Gaussian methods are fast but brittle, that TiNeuVox often wins on quality, and that dataset complexity overwhelms method differences—are summaries of measured PSNR, SSIM, LPIPS, FPS, and training-time numbers computed by running reimplementations of third-party methods on external datasets (D-NeRF, Nerfies, HyperNeRF, NeRF-DS, iPhone) plus a synthetic dataset with controlled camera baseline and object motion. No target quantity is defined in terms of the conclusion: the motion-representation taxonomy is an organizational scheme used to partition results, not a constraint that forces the rank order. The synthetic dataset varies camera baseline and object motion and then the paper reports that performance degrades under those conditions; that is a controlled experiment, not a tautology. The paper explicitly acknowledges the chicken-and-egg difficulty of defining scene complexity independently of method performance in its Limitations section, which is the opposite of smuggling the conclusion into the setup. The assertion that reimplementations 'approximately match the results of the original works' is an unverified fairness/correctness assumption; if false, it would invalidate the benchmark, but it is not a circular derivation because the benchmark numbers are not fitted to the paper's findings. Self-citations (e.g., Stearns et al., Lei et al., Liang et al.) appear only as related work or background and are not load-bearing for the empirical conclusions. No circular step can be exhibited from the paper's own equations or definitions, so the appropriate finding is no significant circularity, score 0.

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

No fitted parameters or invented entities are introduced; the benchmark rests on measurement assumptions about faithful reimplementation, correct masks, accurate camera poses, and metric validity. These assumptions are reasonable but not proven.

assumptions (4)
  • domain assumption Single-codebase reimplementations of the five Gaussian methods faithfully represent the original works, and TiNeuVox is run from its original implementation.
    Invoked in Section 3.3 ('we integrate these algorithms into a single codebase') and asserted in Section 4 ('our implementations approximately match the results of the original works'); all rank-order findings depend on it.
  • domain assumption SAM-Track per-frame binary masks accurately separate dynamic foreground from static background.
    Used in Section 3.2 to compute mPSNR, mSSIM, mMS-SSIM, and mLPIPS; Finding 8 and the mask-based ranking rely on these masks being correct.
  • domain assumption COLMAP camera poses, including the authors' pose corrections for 7/17 HyperNeRF scenes, do not bias comparisons toward or against any motion representation.
    Appendix B.2 shows pose correction improves static rendering but helps some methods (EffGS, STG, RTGS) while hurting others (TiNeuVox, DeformableGS, 4DGS), so pose quality interacts with method identity.
  • domain assumption PSNR, SSIM, MS-SSIM, and LPIPS are accepted measures of view synthesis quality for this comparison.
    Used in Section 3.2 as the evaluation metrics; the conclusions are only as meaningful as these proxies for perceived reconstruction quality.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Monocular Dynamic Gaussian Splatting: Fast, Brittle, and Scene Complexity Rules." pith.science (2026). https://pith.science/paper/3WH7T4TC

@misc{pith2026241204457,
  author       = {Pith},
  title        = {Pith review of: Monocular Dynamic Gaussian Splatting: Fast, Brittle, and Scene Complexity Rules},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3WH7T4TC}},
  note         = {Machine review of arXiv:2412.04457}
}
read the original abstract

Gaussian splatting methods are emerging as a popular approach for converting multi-view image data into scene representations that allow view synthesis. In particular, there is interest in enabling view synthesis for dynamic scenes using only monocular input data -- an ill-posed and challenging problem. The fast pace of work in this area has produced multiple simultaneous papers that claim to work best, which cannot all be true. In this work, we organize, benchmark, and analyze many Gaussian-splatting-based methods, providing apples-to-apples comparisons that prior works have lacked. We use multiple existing datasets and a new instructive synthetic dataset designed to isolate factors that affect reconstruction quality. We systematically categorize Gaussian splatting methods into specific motion representation types and quantify how their differences impact performance. Empirically, we find that their rank order is well-defined in synthetic data, but the complexity of real-world data currently overwhelms the differences. Furthermore, the fast rendering speed of all Gaussian-based methods comes at the cost of brittleness in optimization. We summarize our experiments into a list of findings that can help to further progress in this lively problem setting.

Figures

Figures reproduced from arXiv: 2412.04457 by the authors.

Figure 1
Figure 1. Gaussian splatting with motion. a) Overview of 3D Gaussian Splatting. Adding motion to static 3DGS typically modifies the pipeline only by adding a dynamics module, as marked by the red box. b) Adaptive densification via cloning, splitting, and pruning Gaussians to reflect scene details. This brittle process often causes difficulties during optimization for monocular dynamic sequences. 2 [PITH_FULL_IMAGE:figures/fu… view at source ↗
Figure 2
Figure 2. Iterative methods fail on monocular input. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Left: Illustration of the instructive dataset. A simple textured cube moves in a straight line (distance D) against a differently-textured background. In another condition, the cube additionally performs an overall rotation (ϕ radians). The training-view camera moves in an arc (covering distance B) in the same direction as the cube, tracking it. The test camera (not depicted) travels in a sinusoidal trajectory behin… view at source ↗
Figures from the paper (39 more)
Figure 5
Figure 5. Figure 5: Per-dataset Quantitative Results. Test set LPIPS along with error bars for all methods on each of the different datasets. Note that lower is better. No method is clearly better across datasets [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: # Gaussians after optimization for each method, as a result of the adaptive densification. rendering time, operating at 0.3 FPS compared to 20–200 FPS for the Gaussian methods. The speed difference is due to the rasterization in 3DGS versus volume rendering in NeRF. 4.…
Figure 4
Figure 4. Figure 4: RTGS overfits to training views Motion representation complexity hurts efficiency. Comparing methods on training time and rendering speed, we find that basis-based methods (EffGS & STGs) are faster to train and faster to render than MLP-based methods (De￾formableGS & 4…
Figure 7
Figure 7. Figure 7: Convergence vs. frequency. We see that scenes with higher frequencies take longer to optimize [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Frequency vs. #Gaussians. Scenes with higher frequencies end up with more Gaussians. 8 [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: Render FPS vs. #Gaussians. More Gaussians leads to slower rendering [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]
Figure 10
Figure 10. Figure 10: Train-Test Performance Gaps. We show the difference between the average LPIPS on the train and test set, where a larger gap indicates more overfitting to the training sequence. Note that here larger negative values indicate more severe overfitting [PITH_FULL_IMAGE:fi…
Figure 11
Figure 11. Figure 11: Instability. Training frame and renderings from 3 runs of the field MLP method DeformableGS (Yang et al., 2023). As discussed in Section 2.1, a key detail of Gaussian Splat￾ting methods is adaptive density control: Gaussians are added and removed during optimization, …
Figure 12
Figure 12. Figure 12: Strictly-Monocular Dataset. Test metrics on the strictly-monocular iPhone dataset. spectrum from all images to calculate the mean frequency. Scenes with a predominance of higher frequencies take longer to optimize ( [PITH_FULL_IMAGE:figures/full_fig_p010_12.png]
Figure 13
Figure 13. Figure 13: Specular Dataset. Test results on NeRF-DS, which includes reflective objects [PITH_FULL_IMAGE:figures/full_fig_p011_13.png]
Figure 14
Figure 14. Figure 14: Results on our instructive synthetic dataset. Top Left: LPIPS↓ heatmap shows how camera baseline and object motion range affect performance across all methods on average. Decreasing object motion range (right to left) affects reconstruction performance positively; dec…
Figure 15
Figure 15. Figure 15: Foreground-Only LPIPS (↓) reveals the weakness of 3DGS on dynamic elements. 5.1 Why Motion Representations Matter From an optimization standpoint, any dynamic splatting algorithm must determine where and how Gaussians move as the scene evolves over time. The chosen mo…
Figure 16
Figure 16. Figure 16: DyCheck’s ω sequence diffi￾culty metric vs. PSNR↑/SSIM↑/MS￾SSIM↑/LPIPS↓. Each dot is a sequence from a dataset. There is no strong corre￾lation either within or across datasets. 14 [PITH_FULL_IMAGE:figures/full_fig_p014_16.png]
Figure 17
Figure 17. Figure 17: Per-dataset Quantitative Results. Test set PSNR along with error bars for all methods on each of the datasets. (↑). We see that the datasets have different winning methods [PITH_FULL_IMAGE:figures/full_fig_p021_17.png]
Figure 18
Figure 18. Figure 18: Foreground-Only PSNRs Evaluation (↑) [PITH_FULL_IMAGE:figures/full_fig_p021_18.png]
Figure 19
Figure 19. Figure 19: Train-Test Performance Gaps. We show the difference between the average PSNR on the train and test set, where a larger gap indicates more overfitting to the training sequence. 21 [PITH_FULL_IMAGE:figures/full_fig_p021_19.png]
Figure 20
Figure 20. Figure 20: Per-dataset Quantitative Results. Test set SSIM along with error bars for all methods on each of the datasets. (↑). We see that the datasets have different winning methods [PITH_FULL_IMAGE:figures/full_fig_p022_20.png]
Figure 21
Figure 21. Figure 21: Foreground-Only SSIMs Evaluation (↑) [PITH_FULL_IMAGE:figures/full_fig_p022_21.png]
Figure 22
Figure 22. Figure 22: Train-Test Performance Gaps. We show the difference between the average SSIM on the train and test set, where a larger gap indicates more overfitting to the training sequence. 22 [PITH_FULL_IMAGE:figures/full_fig_p022_22.png]
Figure 23
Figure 23. Figure 23: Per-dataset Quantitative Results. Figure shows the test set MS-SSIM along with error bars for all methods on each of the different datasets. Note that higher is better. We see that the datasets have different winning methods [PITH_FULL_IMAGE:figures/full_fig_p023_23.png]
Figure 24
Figure 24. Figure 24: Foreground-Only MS-SSIMs Evaluation (↑) [PITH_FULL_IMAGE:figures/full_fig_p023_24.png]
Figure 25
Figure 25. Figure 25: Train-Test Performance Gaps. We show the difference between the average MS-SSIM on the train and test set, where a larger gap indicates more overfitting to the training sequence. 23 [PITH_FULL_IMAGE:figures/full_fig_p023_25.png]
Figure 26
Figure 26. Figure 26: Per-dataset Quantitative Results. Figure shows the test set Train Time along with error bars for all methods on each of the different datasets. Note that lower is better. We see that the datasets have different winning methods [PITH_FULL_IMAGE:figures/full_fig_p024_26.png]
Figure 27
Figure 27. Figure 27: Per-dataset Quantitative Results. Figure shows the test set render FPS along with error bars for all methods on each of the different datasets. Note that higher is better. We see that the datasets have different winning methods. A.2.5 Relative Gap vs. Absolute Perform…
Figure 14
Figure 14. Figure 14: fig. 14. We note that 3DGS often ends up higher in the ranking for pixel-difference metrics mPSNR, mSSIM, [PITH_FULL_IMAGE:figures/full_fig_p025_14.png]
Figure 28
Figure 28. Figure 28: Individual ablations for the instructive dataset, measured by LPIPS. The figure shows camera baseline and motion range ablations for each method separately. Top Two Rows: LPIPS↓ heatmaps. Bottom Two Rows: mLPIPS↓ heatmaps. On average the reconstruction becomes harder …
Figure 29
Figure 29. Figure 29: Individual ablations for the instructive dataset, measured by PSNR. The figure shows camera baseline and motion range ablations for each method separately. Top Two Rows: PSNR↑ heatmaps. Bottom Two Rows: mPSNR↑ heatmaps. On average the reconstruction becomes harder wit…
Figure 30
Figure 30. Figure 30: Results on our instructive synthetic dataset. Top Left: PSNR↑ heatmap shows how camera baseline and object motion range affect performance across all methods on average. Increasing object motion range (left to right) affects reconstruction performance negatively; decr…
Figure 31
Figure 31. Figure 31: Individual ablations for the instructive dataset, measured by SSIM. The figure shows camera baseline and motion range ablations for each method separately. Top Two Rows: SSIM↑ heatmaps. Bottom Two Rows: mSSIM↑ heatmaps. On average the reconstruction becomes harder wit…
Figure 32
Figure 32. Figure 32: Results on our instructive synthetic dataset. Top Left: SSIM↑ heatmap shows how camera baseline and object motion range affect performance across all methods on average. Increasing object motion range (left to right) affects reconstruction performance negatively; decr…
Figure 33
Figure 33. Figure 33: Individual ablations for the instructive dataset, measured by MS-SSIM. The figure shows camera baseline and motion range ablations for each method separately. Top Two Rows: MS-SSIM↑ heatmaps. Bottom Two Rows: mMS-SSIM↑ heatmaps. On average the reconstruction becomes h…
Figure 34
Figure 34. Figure 34: Results on our instructive synthetic dataset. Left: MS-SSIM↑ heatmap shows how camera baseline and object motion range affect performance across all methods on average. Increasing object motion range (left to right) affects reconstruction performance negatively; decre…
Figure 35
Figure 35. Figure 35: Comparison of ground-truth and static Gaussian Splatting (Kerbl et al., 2023) results on before [PITH_FULL_IMAGE:figures/full_fig_p033_35.png]
Figure 36
Figure 36. Figure 36: PSNR on the same set of HyperNeRF sequences to show Camera inaccuracy’s effect on recon [PITH_FULL_IMAGE:figures/full_fig_p035_36.png]
Figure 37
Figure 37. Figure 37: SSIM on the same set of HyperNeRF sequences to show Camera inaccuracy’s effect on reconstruction. [PITH_FULL_IMAGE:figures/full_fig_p035_37.png]
Figure 38
Figure 38. Figure 38: MS-SSIM on the same set of HyperNeRF sequences to show Camera inaccuracy’s effect on [PITH_FULL_IMAGE:figures/full_fig_p035_38.png]
Figure 39
Figure 39. Figure 39: LPIPS on the same set of HyperNeRF sequences to show Camera inaccuracy’s effect on recon [PITH_FULL_IMAGE:figures/full_fig_p036_39.png]
Figure 40
Figure 40. Figure 40: Rotational pose noise influence over methods ranking. We observe that although the pose noise is influencing the absolute metrics, the rankings themselves are mostly unchanged. 37 [PITH_FULL_IMAGE:figures/full_fig_p037_40.png]
Figure 41
Figure 41. Figure 41: Qualitative effect of adding rotational noise to the pose. Left: Reconstruction using clean pose. Right: Reconstruction using 5°-distorted poses. Both reconstructions are performed with DeformableGS. 38 [PITH_FULL_IMAGE:figures/full_fig_p038_41.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

64 extracted references · 28 canonical work pages

  1. [1]

    Per-gaussian embedding-based deformation for deformable 3d gaussian splatting, 2024

    Jeongmin Bae, Seoha Kim, Youngsik Yun, Hahyun Lee, Gun Bang, and Youngjung Uh. Per-gaussian embedding-based deformation for deformable 3d gaussian splatting, 2024. URL https://arxiv.org/abs/2404.03613

    arXiv 2024

  2. [2]

    Dyblurf: Dynamic deblurring neural radiance fields for blurry monocular video

    Minh-Quan Viet Bui, Jongmin Park, Jihyong Oh, and Munchurl Kim. Dyblurf: Dynamic deblurring neural radiance fields for blurry monocular video. arXiv preprint arXiv:2312.13528, 2023

    arXiv 2023

  3. [3]

    Hexplane: A fast representation for dynamic scenes

    Ang Cao and Justin Johnson. Hexplane: A fast representation for dynamic scenes. CVPR, 2023

    work page 2023

  4. [4]

    A survey on 3d gaussian splatting, 2024

    Guikun Chen and Wenguan Wang. A survey on 3d gaussian splatting, 2024. URL https://arxiv.org/abs/2401.03890

    arXiv 2024

  5. [5]

    Segment and track anything

    Yangming Cheng, Liulei Li, Yuanyou Xu, Xiaodi Li, Zongxin Yang, Wenguan Wang, and Yi Yang. Segment and track anything. arXiv preprint arXiv:2305.06558, 2023

    arXiv 2023

  6. [6]

    Neural parametric gaussians for monocular non-rigid object reconstruction

    Devikalyan Das, Christopher Wewer, Raza Yunus, Eddy Ilg, and Jan Eric Lenssen. Neural parametric gaussians for monocular non-rigid object reconstruction. arXiv preprint arXiv:2312.01196, 2023

    arXiv 2023

  7. [7]

    Superpoint gaussian splatting for real-time high-fidelity monocular dynamic scene reconstruction

    Gang Zeng Diwen Wan, Ruijie Lu. Superpoint gaussian splatting for real-time high-fidelity monocular dynamic scene reconstruction. In Forty-first International Conference on Machine Learning, 2024

    work page 2024

  8. [8]

    4d gaussian splatting: Towards efficient novel view synthesis for dynamic scenes, 2024

    Yuanxing Duan, Fangyin Wei, Qiyu Dai, Yuhang He, Wenzheng Chen, and Baoquan Chen. 4d gaussian splatting: Towards efficient novel view synthesis for dynamic scenes, 2024. URL https://arxiv.org/abs/2402.03307

    arXiv 2024

Show all 64 references
  1. [9]

    Md-splatting: Learning metric deformation from 4d gaussians in highly deformable scenes

    Bardienus Pieter Duisterhof, Zhao Mandi, Yunchao Yao, Jia-Wei Liu, Mike Zheng Shou, Shuran Song, and Jeffrey Ichnowski. Md-splatting: Learning metric deformation from 4d gaussians in highly deformable scenes. ArXiv, abs/2312.00583, 2023. URL https://api.semanticscholar.org/Cor...

    2023 arXiv

  2. [10]

    Fast dynamic radiance fields with time-aware neural voxels

    Jiemin Fang, Taoran Yi, Xinggang Wang, Lingxi Xie, Xiaopeng Zhang, Wenyu Liu, Matthias Nie ner, and Qi Tian. Fast dynamic radiance fields with time-aware neural voxels. In SIGGRAPH Asia 2022 Conference Papers, pp.\ 1--9, 2022

    2022

  3. [11]

    K-planes: Explicit radiance fields in space, time, and appearance

    Sara Fridovich-Keil, Giacomo Meanti, Frederik Rahbæk Warburg, Benjamin Recht, and Angjoo Kanazawa. K-planes: Explicit radiance fields in space, time, and appearance. In CVPR, 2023

    2023

  4. [12]

    Dynamic view synthesis from dynamic monocular video

    Chen Gao, Ayush Saraf, Johannes Kopf, and Jia-Bin Huang. Dynamic view synthesis from dynamic monocular video. In Proceedings of the IEEE International Conference on Computer Vision, 2021

    2021

  5. [13]

    Dynamic novel-view synthesis: A reality check

    Hang Gao, Ruilong Li, Shubham Tulsiani, Bryan Russell, and Angjoo Kanazawa. Dynamic novel-view synthesis: A reality check. In NeurIPS, 2022

    2022

  6. [14]

    Gaussianflow: Splatting gaussian dynamics for 4d content creation, 2024

    Quankai Gao, Qiangeng Xu, Zhe Cao, Ben Mildenhall, Wenchao Ma, Le Chen, Danhang Tang, and Ulrich Neumann. Gaussianflow: Splatting gaussian dynamics for 4d content creation, 2024. URL https://arxiv.org/abs/2403.12365

    2024 arXiv

  7. [15]

    Motion-aware 3d gaussian splatting for efficient dynamic scene reconstruction, 2024

    Zhiyang Guo, Wengang Zhou, Li Li, Min Wang, and Houqiang Li. Motion-aware 3d gaussian splatting for efficient dynamic scene reconstruction, 2024. URL https://arxiv.org/abs/2403.11447

    2024 arXiv

  8. [16]

    Sc-gs: Sparse-controlled gaussian splatting for editable dynamic scenes

    Yi-Hua Huang, Yang-Tian Sun, Ziyi Yang, Xiaoyang Lyu, Yan-Pei Cao, and Xiaojuan Qi. Sc-gs: Sparse-controlled gaussian splatting for editable dynamic scenes. arXiv preprint arXiv:2312.14937, 2023

    2023 arXiv

  9. [17]

    D-tensorf: Tensorial radiance fields for dynamic scenes

    Hankyu Jang and Daeyoung Kim. D-tensorf: Tensorial radiance fields for dynamic scenes. ArXiv, abs/2212.02375, 2022. URL https://api.semanticscholar.org/CorpusID:254247189

    2022 arXiv

  10. [18]

    Johnson, Marc Habermann, Soshi Shimada, Vladislav Golyanik, and Christian Theobalt

    Erik C.M. Johnson, Marc Habermann, Soshi Shimada, Vladislav Golyanik, and Christian Theobalt. Unbiased 4d: Monocular 4d reconstruction with a neural deformation model. CVPR Workshop, 2023

    2023

  11. [19]

    Map visibility estimation for large-scale dynamic 3d reconstruction

    Hanbyul Joo, Hyun Soo Park, and Yaser Sheikh. Map visibility estimation for large-scale dynamic 3d reconstruction. In IEEE Conference on Computer Vision and Pattern Recognition, 2014

    2014

  12. [20]

    An efficient 3d gaussian representation for monocular/multi-view dynamic scenes, 2023

    Kai Katsumata, Duc Minh Vo, and Hideki Nakayama. An efficient 3d gaussian representation for monocular/multi-view dynamic scenes, 2023. URL https://arxiv.org/abs/2311.12897

    2023 arXiv

  13. [21]

    3d gaussian splatting for real-time radiance field rendering

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk \"u hler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering. ACM Transactions on Graphics, 42 0 (4), July 2023. URL https://repo-sam.inria.fr/fungraph/3d-gaussian-splatting/

    2023

  14. [22]

    Approximate differentiable rendering with algebraic surfaces

    Leonid Keselman and Martial Hebert. Approximate differentiable rendering with algebraic surfaces. In European Conference on Computer Vision (ECCV), 2022

    2022

  15. [23]

    Nersemble: Multi-view radiance field reconstruction of human heads

    Tobias Kirschstein, Shenhan Qian, Simon Giebenhain, Tim Walter, and Matthias Nie ner. Nersemble: Multi-view radiance field reconstruction of human heads. ACM Trans. Graph., 42 0 (4), jul 2023. ISSN 0730-0301. doi:10.1145/3592455. URL https://doi.org/10.1145/3592455

    2023 doi

  16. [24]

    Dynmf: Neural motion factorization for real-time dynamic view synthesis with 3d gaussian splatting

    Agelos Kratimenos, Jiahui Lei, and Kostas Daniilidis. Dynmf: Neural motion factorization for real-time dynamic view synthesis with 3d gaussian splatting. arXiV, 2023

    2023

  17. [25]

    Mosca: Dynamic gaussian fusion from casual videos via 4d motion scaffolds

    Jiahui Lei, Yijia Weng, Adam Harley, Leonidas Guibas, and Kostas Daniilidis. Mosca: Dynamic gaussian fusion from casual videos via 4d motion scaffolds. arXiv preprint arXiv:2405.17421, 2024

    2024 arXiv

  18. [26]

    Neural 3d video synthesis from multi-view video

    Tianye Li, Mira Slavcheva, Michael Zollhoefer, Simon Green, Christoph Lassner, Changil Kim, Tanner Schmidt, Steven Lovegrove, Michael Goesele, Richard Newcombe, et al. Neural 3d video synthesis from multi-view video. In Proceedings of the IEEE/CVF Conference on Computer Vision...

    2022

  19. [27]

    Spacetime gaussian feature splatting for real-time dynamic view synthesis

    Zhan Li, Zhang Chen, Zhong Li, and Yi Xu. Spacetime gaussian feature splatting for real-time dynamic view synthesis. arXiv preprint arXiv:2312.16812, 2023 a

    2023 arXiv

  20. [28]

    Neural scene flow fields for space-time view synthesis of dynamic scenes

    Zhengqi Li, Simon Niklaus, Noah Snavely, and Oliver Wang. Neural scene flow fields for space-time view synthesis of dynamic scenes. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021

    2021

  21. [29]

    Dynibar: Neural dynamic image-based rendering

    Zhengqi Li, Qianqian Wang, Forrester Cole, Richard Tucker, and Noah Snavely. Dynibar: Neural dynamic image-based rendering. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023 b

    2023

  22. [30]

    Gaufre: Gaussian deformation fields for real-time dynamic novel view synthesis, 2023

    Yiqing Liang, Numair Khan, Zhengqin Li, Thu Nguyen-Phuoc, Douglas Lanman, James Tompkin, and Lei Xiao. Gaufre: Gaussian deformation fields for real-time dynamic novel view synthesis, 2023. URL https://arxiv.org/abs/2312.11458

    2023 arXiv

  23. [31]

    Gaussian-flow: 4d reconstruction with dynamic 3d gaussian particle

    Youtian Lin, Zuozhuo Dai, Siyu Zhu, and Yao Yao. Gaussian-flow: 4d reconstruction with dynamic 3d gaussian particle. arXiv:2312.03431, 2023

    2023 arXiv

  24. [32]

    Modgs: Dynamic gaussian splatting from causually-captured monocular videos, 2024

    Qingming Liu, Yuan Liu, Jiepeng Wang, Xianqiang Lv, Peng Wang, Wenping Wang, and Junhui Hou. Modgs: Dynamic gaussian splatting from causually-captured monocular videos, 2024. URL https://arxiv.org/abs/2406.00434

    2024 arXiv

  25. [33]

    Robust dynamic radiance fields

    Yu-Lun Liu, Chen Gao, Andreas Meuleman, Hung-Yu Tseng, Ayush Saraf, Changil Kim, Yung-Yu Chuang, Johannes Kopf, and Jia-Bin Huang. Robust dynamic radiance fields. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023

    2023

  26. [34]

    3d geometry-aware deformable gaussian splatting for dynamic view synthesis

    Zhicheng Lu, Xiang Guo, Le Hui, Tianrui Chen, Ming Yang, Xiao Tang, Feng Zhu, and Yuchao Dai. 3d geometry-aware deformable gaussian splatting for dynamic view synthesis. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024

    2024

  27. [35]

    Dynamic 3d gaussians: Tracking by persistent dynamic view synthesis, 2023

    Jonathon Luiten, Georgios Kopanas, Bastian Leibe, and Deva Ramanan. Dynamic 3d gaussians: Tracking by persistent dynamic view synthesis, 2023. URL https://arxiv.org/abs/2308.09713

    2023 arXiv

  28. [36]

    Srinivasan, Matthew Tancik, Jonathan T

    Ben Mildenhall, Pratul P. Srinivasan, Matthew Tancik, Jonathan T. Barron, Ravi Ramamoorthi, and Ren Ng. Nerf: Representing scenes as neural radiance fields for view synthesis. In ECCV, 2020

    2020

  29. [37]

    Newcombe, Dieter Fox, and Steven M

    Richard A. Newcombe, Dieter Fox, and Steven M. Seitz. Dynamicfusion: Reconstruction and tracking of non-rigid scenes in real-time. In 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp.\ 343--352, 2015. doi:10.1109/CVPR.2015.7298631

    2015

  30. [38]

    Barron, Sofien Bouaziz, Dan B Goldman, Steven M

    Keunhong Park, Utkarsh Sinha, Jonathan T. Barron, Sofien Bouaziz, Dan B Goldman, Steven M. Seitz, and Ricardo Martin-Brualla. Nerfies: Deformable neural radiance fields. ICCV, 2021 a

    2021

  31. [39]

    Barron, Sofien Bouaziz, Dan B Goldman, Ricardo Martin-Brualla, and Steven M

    Keunhong Park, Utkarsh Sinha, Peter Hedman, Jonathan T. Barron, Sofien Bouaziz, Dan B Goldman, Ricardo Martin-Brualla, and Steven M. Seitz. Hypernerf: A higher-dimensional representation for topologically varying neural radiance fields. ACM Trans. Graph., 40 0 (6), dec 2021 b

    2021

  32. [40]

    D-nerf: Neural radiance fields for dynamic scenes

    Albert Pumarola, Enric Corona, Gerard Pons-Moll, and Francesc Moreno-Noguer. D-nerf: Neural radiance fields for dynamic scenes. arXiv preprint arXiv:2011.13961, 2020

    2011 arXiv

  33. [41]

    Blirf: Band limited radiance fields for dynamic scene modeling

    Sameera Ramasinghe, Violetta Shevchenko, Gil Avraham, and Anton van den Hengel. Blirf: Band limited radiance fields for dynamic scene modeling. In AAAI 2024, 2024. URL https://www.amazon.science/publications/blirf-band-limited-radiance-fields-for-dynamic-scene-modeling

    2024

  34. [42]

    Structure-from-motion revisited

    Johannes Lutz Sch\" o nberger and Jan-Michael Frahm. Structure-from-motion revisited. In Conference on Computer Vision and Pattern Recognition (CVPR), 2016

    2016

  35. [43]

    Killingfusion: Non-rigid 3d reconstruction without correspondences

    Miroslava Slavcheva, Maximilian Baust, Daniel Cremers, and Slobodan Ilic. Killingfusion: Non-rigid 3d reconstruction without correspondences. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp.\ 5474--5483, 2017. URL https://api.semanticscholar.org/Corp...

    2017

  36. [44]

    Nerfplayer: A streamable dynamic scene representation with decomposed neural radiance fields

    Liangchen Song, Anpei Chen, Zhong Li, Zhang Chen, Lele Chen, Junsong Yuan, Yi Xu, and Andreas Geiger. Nerfplayer: A streamable dynamic scene representation with decomposed neural radiance fields. IEEE Transactions on Visualization and Computer Graphics, 29 0 (5): 0 2732--2742,...

    2023

  37. [45]

    Dynamic gaussian marbles for novel view synthesis of casual monocular videos

    Colton Stearns, Adam W Harley, Mikaela Uy, Florian Dubost, Federico Tombari, Gordon Wetzstein, and Leonidas Guibas. Dynamic gaussian marbles for novel view synthesis of casual monocular videos. In SIGGRAPH Asia 2024 Conference Papers, pp.\ 1--11, 2024

    2024

  38. [46]

    3dgstream: On-the-fly training of 3d gaussians for efficient streaming of photo-realistic free-viewpoint videos

    Jiakai Sun, Han Jiao, Guangyuan Li, Zhanjie Zhang, Lei Zhao, and Wei Xing. 3dgstream: On-the-fly training of 3d gaussians for efficient streaming of photo-realistic free-viewpoint videos. arXiv preprint arXiv:2403.01444, 2024

    2024 arXiv

  39. [47]

    Raft: Recurrent all-pairs field transforms for optical flow

    Zachary Teed and Jia Deng. Raft: Recurrent all-pairs field transforms for optical flow. In European Conference on Computer Vision, 2020. URL https://api.semanticscholar.org/CorpusID:214667893

    2020

  40. [48]

    Non-rigid neural radiance fields: Reconstruction and novel view synthesis of a dynamic scene from monocular video, 2020

    Edgar Tretschk, Ayush Tewari, Vladislav Golyanik, Michael Zollhöfer, Christoph Lassner, and Christian Theobalt. Non-rigid neural radiance fields: Reconstruction and novel view synthesis of a dynamic scene from monocular video, 2020. URL https://arxiv.org/abs/2012.12247

    2020 arXiv

  41. [49]

    Neural trajectory fields for dynamic novel view synthesis

    Chaoyang Wang, Ben Eckart, Simon Lucey, and Orazio Gallo. Neural trajectory fields for dynamic novel view synthesis. arXiv preprint arXiv:2105.05994, 2021

    2021 arXiv

  42. [50]

    Jeni, and Simon Lucey

    Chaoyang Wang, Lachlan Ewen MacDonald, L\'aszl\'o A. Jeni, and Simon Lucey. Flow supervision for deformable nerf. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp.\ 21128--21137, June 2023

    2023

  43. [51]

    Shape of motion: 4d reconstruction from a single video, 2024

    Qianqian Wang, Vickie Ye, Hang Gao, Jake Austin, Zhengqi Li, and Angjoo Kanazawa. Shape of motion: 4d reconstruction from a single video, 2024. URL https://arxiv.org/abs/2407.13764

    2024

  44. [52]

    Multiscale structural similarity for image quality assessment

    Zhou Wang, Eero P Simoncelli, and Alan C Bovik. Multiscale structural similarity for image quality assessment. In The Thrity-Seventh Asilomar Conference on Signals, Systems & Computers, 2003, volume 2, pp.\ 1398--1402. Ieee, 2003

    2003

  45. [53]

    Image quality assessment: from error visibility to structural similarity

    Zhou Wang, Alan C Bovik, Hamid R Sheikh, and Eero P Simoncelli. Image quality assessment: from error visibility to structural similarity. IEEE transactions on image processing, 13 0 (4): 0 600--612, 2004

    2004

  46. [54]

    4d gaussian splatting for real-time dynamic scene rendering

    Guanjun Wu, Taoran Yi, Jiemin Fang, Lingxi Xie, Xiaopeng Zhang, Wei Wei, Wenyu Liu, Qi Tian, and Wang Xinggang. 4d gaussian splatting for real-time dynamic scene rendering. arXiv preprint arXiv:2310.08528, 2023

    2023 arXiv

  47. [55]

    Bridging 3d gaussian and mesh for freeview video rendering, 2024

    Yuting Xiao, Xuan Wang, Jiafei Li, Hongrui Cai, Yanbo Fan, Nan Xue, Minghui Yang, Yujun Shen, and Shenghua Gao. Bridging 3d gaussian and mesh for freeview video rendering, 2024. URL https://arxiv.org/abs/2403.11453

    2024 arXiv

  48. [56]

    Neural fields in visual computing and beyond

    Yiheng Xie, Towaki Takikawa, Shunsuke Saito, Or Litany, Shiqin Yan, Numair Khan, Federico Tombari, James Tompkin, Vincent Sitzmann, and Srinath Sridhar. Neural fields in visual computing and beyond. Computer Graphics Forum, 2022. ISSN 1467-8659. doi:10.1111/cgf.14505

    2022 doi

  49. [57]

    Nerf-ds: Neural radiance fields for dynamic specular objects

    Zhiwen Yan, Chen Li, and Gim Hee Lee. Nerf-ds: Neural radiance fields for dynamic specular objects. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp.\ 8285--8295, 2023

    2023

  50. [58]

    Real-time photorealistic dynamic scene representation and rendering with 4d gaussian splatting

    Zeyu Yang, Hongye Yang, Zijie Pan, and Li Zhang. Real-time photorealistic dynamic scene representation and rendering with 4d gaussian splatting. In International Conference on Learning Representations (ICLR), 2024

    2024

  51. [59]

    Deformable 3d gaussians for high-fidelity monocular dynamic scene reconstruction

    Ziyi Yang, Xinyu Gao, Wen Zhou, Shaohui Jiao, Yuqing Zhang, and Xiaogang Jin. Deformable 3d gaussians for high-fidelity monocular dynamic scene reconstruction. arXiv preprint arXiv:2309.13101, 2023

    2023 arXiv

  52. [60]

    Novel view synthesis of dynamic scenes with globally coherent depths from a monocular camera

    Jae Shin Yoon, Kihwan Kim, Orazio Gallo, Hyun Soo Park, and Jan Kautz. Novel view synthesis of dynamic scenes with globally coherent depths from a monocular camera. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp.\ 5335--5344, 2020. URL https://a...

    2020

  53. [61]

    Cogs: Controllable gaussian splatting

    Heng Yu, Joel Julin, Zoltan A Milacski, Koichiro Niinuma, and Laszlo A Jeni. Cogs: Controllable gaussian splatting. arXiv, 2023

    2023

  54. [62]

    The unreasonable effectiveness of deep features as a perceptual metric

    Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shechtman, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. In CVPR, 2018

    2018

  55. [63]

    Surface splatting

    Matthias Zwicker, Hanspeter Pfister, Jeroen Van Baar, and Markus Gross. Surface splatting. In Proceedings of the 28th annual conference on Computer graphics and interactive techniques, pp.\ 371--378, 2001

    2001

  56. [64]

    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.