Pith. sign in

REVIEW 3 major objections 5 minor 42 references

XClipGS: Exact Half-Space Clipping for Medical Volume Gaussian Splatting

T0 review · 3 major / 5 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 paper shows that clipping a Gaussian-splatted medical volume at an arbitrary plane reduces to a closed-form per-pixel Gaussian-CDF factor, making interactive cut views exact under the affine EWA model and free of learned clipping…

desk verdict A clean analytic clipping operator for EWA splatting, honestly scoped; the unquantified affine-vs-perspective gap and single-seed numbers are the main things to push on in review. read the letter →

arxiv 2608.07760 v1 pith:WQB47T6M submitted 2026-08-07 cs.CV cs.GR

Zhongpai Gao , Benjamin Planche , Meng Zheng , Anwesa Choudhuri , Chaoyi Zhou , Terrence Chen , Ziyan Wu This is my paper · ORCID
classification cs.CVcs.GR
keywords Gaussiansplattingvolumerenderingmedicalimagingclippingplanehalf-spaceEWAselection-normaldistributioninteractive
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

The paper claims that when a clipping plane cuts through a medical volume represented by Gaussian splats, the ray integral of a half-space-restricted Gaussian factorizes exactly into the ordinary 2D footprint times a conditional Gaussian CDF whose argument is affine in pixel coordinates. This closed-form per-pixel clip operator is differentiable and adds no learned parameters, so a viewer can move a clip plane through a CT or MRI volume in real time while the exposed interior is learned from clipped reference views. The paper supports the claim with a derivation under the local affine projection model of EWA splatting, and with experiments on eight CT and MRI volumes showing the highest PSNR and the least culled-side leakage among the compared methods.

What carries the argument

The core object is the conditional law of the clip coordinate $\omega = n^\top x$ given the pixel position under affine projection. Its conditional variance is the Schur complement $s^2 = \sigma_n^2 - b^\top A^{-1}b$, and its conditional mean produces the affine CDF argument $\ell(\delta) = k^\top \delta + h$. The per-pixel factor $\alpha K_G(\delta;A)\Phi(\ell(\delta))$ is the selection-normal form for the truncated density, and it reduces clipping to three transient per-primitive coefficients plus one CDF evaluation per covered sample, while remaining differentiable with respect to both the primitive parameters and the clipping plane.

What would settle it

Take one Gaussian primitive that straddles a known plane, render it under the affine EWA model, and compare Equation (5) pixel by pixel with a dense numerical integration of the truncated 3D density along the affine viewing rays. Agreement to floating-point tolerance confirms the identity; any systematic disagreement would falsify the factorization. Repeating under a true perspective camera with large primitives should show divergence, which would confirm the stated affine scope rather than the claim's failure.

Watch

Extended reading notes

Core claim

The paper's central claim is that, under the local affine projection model of EWA splatting, the per-pixel contribution of a Gaussian primitive restricted to a half-space remains closed form: it is the ordinary unclipped footprint $\alpha K_G(\delta; A)$ times a conditional Gaussian CDF $\Phi(k^\top \delta + h)$, with $k$ and $h$ affine in the pixel offset $\delta$. Proposition 1 states that this expression exactly equals the viewing-ray integral of the unnormalized half-space-restricted Gaussian density. The paired screen position and clip coordinate are jointly Gaussian under the affine model, so conditioning on the pixel factorizes the truncated ray integral; the CDF then locates the cut inside each splat rather than attenuating the whole primitive uniformly.

Load-bearing premise

The load-bearing premise is that projection is locally affine, so screen position and clip coordinate are jointly Gaussian; with true perspective projection and large primitives this is an approximation, and Proposition 1 inherits that approximation.

Editorial extensions

If this is right

  • A medical volume proxy trained once with clip-aware views can be cut at any plane axis or orientation at render time, with no plane-conditioned network and no retraining.
  • The analytic operator makes the exposed cut face differentiable, so interior supervision reduces to ordinary photometric training; removing the clipped training views costs 8.3 dB on held-out clipped views.
  • Because the clipped contribution is the unclipped footprint times a CDF factor, render cost stays near hard-cull speed: the paper measures over 650 FPS at 1600x1600 resolution.
  • Localized paired clipped/unclipped metrics (band SSIM, difference-referenced cut error, culled-side leakage) expose cut-face quality that global PSNR and SSIM dilute, and the method leads all four on both voxel-axis and arbitrary-normal planes.

Reading between the lines

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

  • We infer that the same Gaussian-CDF factor could apply to any half-space restriction of a Gaussian splat, such as editable segmentations or multi-plane dissection, since the derivation depends only on the affine projection and the plane equation.
  • We infer that the operator's behavior as the plane approaches the viewing-ray direction (the conditional-variance limit $s \to 0$) gives it a natural advantage in grazing-angle cuts; a targeted study comparing numerical integration at grazing angles could verify this directly.
  • We infer that the paired clipped/unclipped evaluation with difference-referenced error could be reused as a standard protocol for clip operators, separating removal accuracy from reconstruction error in any volume renderer.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes XClipGS, a clipping operator for Gaussian-splatting volume rendering. Under the local affine projection model of EWA splatting (Eq. 1), the authors derive a closed-form per-pixel factor for a plane-truncated Gaussian: the clipped contribution equals the unclipped EWA footprint times Phi(k^T delta + h), where k and h are computed from the primitive covariance, plane normal/offset, and projection Jacobian (Eq. 5). Proposition 1 (proved in Supplement S1) states that this factor equals the viewing-ray integral of the half-space-restricted Gaussian density, up to the per-splat normalization shared with the unclipped footprint. The operator has no learned parameters and is differentiable. The paper couples this with clip-aware interior supervision using multi-distance views with varied plane axes and offsets (Eq. 6), and introduces cut-face metrics CDE, Leak, and a geometric diagnostic CErr3D. On eight CT and MRI volumes with held-out plane offsets and arbitrary-normal planes, the method reports the best average on all reported metrics, over 650 FPS, and reduced leakage relative to a ClipGS reimplementation, moment-matched (MM) and hard-cull (HC) surrogates, and RaRa.

Significance. If the claims hold, the paper provides the first closed-form, parameter-free per-pixel clipping rule for Gaussian splatting: an exact factorization of the half-space-restricted Gaussian ray integral within the affine EWA model, cleanly separating the render-time operator from interior supervision. The manuscript has notable strengths: the derivation is first-principles and the proof in S1 is careful; the CUDA forward and backward passes are validated against a float64 autograd implementation to roughly 1e-6 relative error (S2); the experimental protocol includes controlled operator swaps on fixed interiors, a supervision ablation, numerically held-out offsets, arbitrary-normal transfer without retraining, and metric-parameter sensitivity sweeps; and the zero-by-construction CErr3D diagnostic is honestly distinguished from image-space metrics. The reported margins are large (about 1.2 dB PSNR over the ClipGS reimplementation, 0.05 band SSIM, roughly 40x leakage reduction on the voxel-axis set), which supports practical relevance if the numbers hold.

major comments (3)
  1. [§3.1–§3.2, Prop. 1; Supp. S2] The exactness claim is established and validated only within the affine EWA model of Eq. (1). Supplement S2's float64 validation compares the CUDA rasterizer against an autograd implementation of the same affine formulas, so it bounds implementation error but cannot bound the divergence between the affine-approximation rays and the true perspective rays of the reference renderer. Because the title and the practical reading of "exact" concern the rendered image, the manuscript should quantify this gap in the reported configuration: for example, measure the per-primitive screen-space residual of Eq. (1) over the projected Gaussian support at the training and test camera distances (the checkpoints are available), or compare the analytic operator's output against a perspective-ray integration of the same restricted Gaussian on a subset of cut-eval views, and report the effect on Leak and Sb. If the gap is non-negligible at the close-up distances used in Table S5 and at the 60-degree FOV, the abstract's "exact" framing should be tempered; if it is small, the numbers should be stated so readers can see the boundary of the claim.
  2. [Supp. S3, Optimization] All empirical claims rest on one training run per scan–method pair at seed 0 (Supplement S3, "Optimization"). Densification and plane sampling are stochastic, so run-to-run variance is unmeasured. The headline margins are large, but the per-volume PSNR leads (Table 1) and the cut-face gaps (Table 2) would be more convincing with at least one additional seed reported as mean and range for the main comparisons, or with an explicit argument for why the fixed-seed pipeline is representative.
  3. [Supp. S4; Table 1] The headline system-level comparison (33.56 versus 32.34 dB, "highest PSNR on every volume") is measured against a best-effort reimplementation of ClipGS from its published description, since no public code was available (Supp. S4). The controlled MM/HC ladder does not depend on this baseline, but the abstract's headline contrast does. I ask the authors to release the reimplementation code, or to state clearly which claims would survive a comparison with the authors' original implementation, and to keep the controlled same-backbone ablations as the primary evidence for the operator's effect.
minor comments (5)
  1. [§4.1] The statement that band PSNR "differs by only hundredths and is omitted" is surprising given the 0.05 average gap in band SSIM between Ours and ClipGS in Table 2; a supporting sentence or a small table in Supplement S5 would help readers understand why SSIM moves while PSNR does not.
  2. [Supp. S5, near-edge line proxy] The grazing-view metric uses the least-squares line proxy l = pinv(P)^T [n; -tau] because a 3D plane projects nonlinearly under perspective; please report the pixel error of this proxy against the true projected plane for the 1.5-degree grazing cameras, so that the CDEg values can be read as metric noise rather than operator error.
  3. [Supp. S4, RaRa integration] The three compatibility fixes to RaRa's released kernel change baseline behavior for fully visible primitives and for chords with both endpoints invisible; documenting the effect of each fix (for example, reporting Leak and CDE for the unfixed kernel) would make the 16x leakage comparison fully auditable.
  4. [Supp. S3, Backbones and warm start] The ClipGS system baseline uses vanilla 3DGS without the Mip-Splatting filter while Ours/MM/HC use the filter; since filtering affects boundary appearance, this additional confound should be acknowledged in §4.2 alongside the representation difference.
  5. [§3.2, Eq. (3)] The vector d_n = Sigma n / sigma_n is introduced only in the prose around Eq. (3); defining it inside the equation block would improve readability.

Circularity Check

0 steps flagged · score 2.0 of 10

No load-bearing circularity: the clip operator is derived from first principles under affine EWA, and evaluation uses held-out offsets and arbitrary normals; remaining self-citations are non-central.

full rationale

The central derivation (Proposition 1, Eq. 5) is self-contained: it conditions the jointly Gaussian screen/clip coordinates induced by Eq. (1), computes the Schur-complement variance (Eq. 4), and obtains the Gaussian-CDF factor; the supplement's S1 proof carries this out explicitly and does not invoke the paper's own prior work. The affine-EWA assumption is inherited from the external EWA splatting literature (Zwicker et al. 2001; 3DGS), and the paper explicitly scopes exactness to that model and notes that perspective-ray methods require re-derived coefficients (Section 3.2, citing 3DGEER). This is a stated modeling limitation, not a circular step. The N-DGS backbone and Render-FM initialization are self-citations (Gao et al. 2025a,b; 2026), but they supply the base representation and warm start only; the clip operator introduces no learned clipping parameters and its analytical form does not depend on those works. Training uses clipped reference views and evaluation uses held-out plane offsets and arbitrary orientations absent from training, so the comparison is not forced. The CErr3D diagnostic is zero by construction for Ours, but the paper explicitly identifies it as a by-construction diagnostic and relies on the independent path-traced reference for image-space Leak; this is disclosed, not circular. The remaining concern, unquantified divergence between affine rays and true perspective rays, is a correctness and applicability risk, not circularity.

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

The analytic clip operator adds no fitted parameters and no invented entities. It relies on the affine EWA projection model, standard Gaussian conditioning, and an external path-traced reference renderer. Remaining hand-set choices are training and evaluation hyperparameters, not fitted to the target result.

free parameters (3)
  • lambda_s (D-SSIM weight in Eq. 6) = 0.2
    Hand-chosen for all experiments; controls the balance between L1 and D-SSIM loss but does not enter the clipping derivation.
  • Retained-mass cutoff and variance floor = Phi(a) <= 1e-6, s^2 >= 1e-8
    Numerical safeguards in the CUDA rasterizer (Supplement S2); hand-set, not fitted to the target result.
  • Metric thresholds for CDE and Leak = theta_fg = 0.04, delta = 2 px, w = 60 px
    Hand-chosen defaults for the cut-face metrics; Supplement S5 tests sensitivity and finds the leakage and face-CDE conclusions stable.
assumptions (4)
  • domain assumption Local affine EWA projection model: u(x) ≈ p_hat_i + M_i(x - mu_i) with M_i = J_i W (Eq. 1).
    Underlies the joint Gaussianity of screen and clip coordinates; Proposition 1 is exact only within this model.
  • domain assumption The physically based path tracer with volumetric half-space clipping provides correct clipped ground truth.
    All supervision and evaluation rely on this external oracle; proxy fidelity is measured against it.
  • standard math Standard Gaussian conditioning and selection-normal formulas (Arellano-Valle et al. 2006; Tallis 1961).
    Used to derive the conditional CDF in Eq. 5 and the moment-matched truncation in Eq. 3.
  • domain assumption Front-to-back alpha compositing and splat order are retained unchanged from the base rasterizer.
    The exactness claim is per-primitive and inherits these renderer choices; image-level output is not globally exact.

how reviews work

0 comments
Cite this review

Pith. "Pith review of XClipGS: Exact Half-Space Clipping for Medical Volume Gaussian Splatting." pith.science (2026). https://pith.science/paper/WQB47T6M

@misc{pith2026260807760,
  author       = {Pith},
  title        = {Pith review of: XClipGS: Exact Half-Space Clipping for Medical Volume Gaussian Splatting},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WQB47T6M}},
  note         = {Machine review of arXiv:2608.07760}
}
read the original abstract

Gaussian-splatting proxies enable interactive rendering of volumetric medical scans, but a clipping plane exposes anatomy not constrained by external-view training and intersects primitives that conventional splatting can only keep or drop whole. We present XClipGS (eXact Clipping), which treats these as two separate problems: the render-time clip operator and supervision of the hidden interior. Under the local affine model used by EWA splatting, the ray integral of a half-space-restricted Gaussian factorizes exactly into its ordinary 2D footprint and a conditional Gaussian CDF whose argument is affine in pixel coordinates. The resulting closed-form per-pixel operator introduces no learned clipping parameters or auxiliary network and remains differentiable with respect to the primitive and plane. We use multi-distance reference views with varied clipping-plane axes and offsets to supervise the interior through the same operator. We also introduce a paired clipped/unclipped cut-face protocol with difference-referenced cut error (CDE) and culled-side leakage (Leak), because global image metrics dilute errors near the plane. On eight CT and MRI volumes with plane offsets not used for training, XClipGS attains the highest PSNR on every volume (33.56 versus 32.34 dB for ClipGS) while rendering at over 650 FPS, far above real time, versus 278 FPS. On voxel-axis cut-face views, it raises average band SSIM from 0.809 to 0.860 and leaks roughly 40 times less. Without retraining, it also achieves the best average across all four metrics on arbitrary-normal planes; on a fixed interior, it matches RaRa's face fidelity with about 16 times less leakage. Project page: https://gaozhongpai.github.io/XClipGS/

Figures

Figures reproduced from arXiv: 2608.07760 by the authors.

Figure 1
Figure 1. Faithful interactive clipping. XClipGS, reference volume rendering, and ClipGS (Li et al. 2025a) (left to right) on a held-out test cut. The arrow marks ClipGS leakage; labels report quality and speed. Abstract Gaussian-splatting proxies enable interactive rendering of vol￾umetric medical scans, but a clipping plane exposes anatomy not constrained by external-view training and intersects prim￾itives that conventiona… view at source ↗
Figure 2
Figure 2. XClipGS pipeline. Exact per-splat half-space integration clips straddling Gaussians, while multi-distance cut views supervise the hidden interior through the same differentiable operator. Why the factor is per pixel. The marginal kept-mass frac￾tion Φ(a) from Equation 2 is a single scalar for the whole primitive. Multiplying the projected footprint uniformly by this value would attenuate the splat but could not loca… view at source ↗
Figure 3
Figure 3. Held-out qualitative comparison. Full views locate each crop; the middle row is an unseen arbitrary-normal grazing view. The remaining columns compare matched reference and method crops, with SSIM and PSNR shown below. Global metrics nevertheless dilute the operator effect, be￾cause half of the held-out views contain no cut and, in the remaining views, the boundary occupies only a small image region. The largest Our… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 37 canonical work pages

  1. [1]

    B.; Branco, M

    Arellano-Valle, R. B.; Branco, M. D.; and Genton, M. G. 2006. A Unified View on Skewed Distributions Arising from Selections. Canadian Journal of Statistics, 34(4): 581--601

    work page 2006

  2. [2]

    T.; Mildenhall, B.; Tancik, M.; Hedman, P.; Martin-Brualla, R.; and Srinivasan, P

    Barron, J. T.; Mildenhall, B.; Tancik, M.; Hedman, P.; Martin-Brualla, R.; and Srinivasan, P. P. 2021. Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance Fields. In ICCV, 5855--5864

    work page 2021

  3. [3]

    Cai, Y.; Liang, Y.; Wang, J.; Wang, A.; Zhang, Y.; Yang, X.; Zhou, Z.; and Yuille, A. 2024. Radiative G aussian Splatting for Efficient X -ray Novel View Synthesis. In ECCV

    work page 2024

  4. [4]

    Cen, J.; Fang, J.; Yang, C.; Xie, L.; Zhang, X.; Shen, W.; and Tian, Q. 2025. Segment Any 3D G aussians. In AAAI

    work page 2025

  5. [5]

    Chen, Y.; Chen, Z.; Zhang, C.; Wang, F.; Yang, X.; Wang, Y.; Cai, Z.; Yang, L.; Liu, H.; and Lin, G. 2024. GaussianEditor : Swift and Controllable 3D Editing with G aussian Splatting. In CVPR

    work page 2024

  6. [6]

    Comaniciu, D.; Engel, K.; Georgescu, B.; and Mansi, T. 2016. Shaping the future through innovations: From medical imaging to precision medicine. Medical Image Analysis, 33: 19--26

    work page 2016

  7. [7]

    Dappa, E.; Higashigaito, K.; Fornaro, J.; Leschka, S.; Wildermuth, S.; and Alkadhi, H. 2016. Cinematic rendering -- an alternative to volume rendering for 3D computed tomography imaging. Insights into Imaging, 7(6): 849--856

    work page 2016

  8. [8]

    R.; and Hu, S.-M

    Feng, Q.-Y.; Cao, G.-C.; Chen, H.-X.; Xu, Q.-C.; Mu, T.-J.; Martin, R. R.; and Hu, S.-M. 2024. EVSplitting : An Efficient and Visually Consistent Splitting Algorithm for 3D G aussian Splatting. In SIGGRAPH Asia 2024 Conference Papers, 1--11

    work page 2024

Show all 42 references
  1. [9]

    Gao, Z.; Planche, B.; Zheng, M.; Chen, X.; Chen, T.; and Wu, Z. 2024. DDGS-CT: Direction-Disentangled Gaussian Splatting for Realistic Volume Rendering. Advances in Neural Information Processing Systems

    2024

  2. [10]

    Gao, Z.; Planche, B.; Zheng, M.; Choudhuri, A.; Chen, T.; and Wu, Z. 2025 a . 6dgs: Enhanced direction-aware gaussian splatting for volumetric rendering. In ICLR

    2025

  3. [11]

    Gao, Z.; Planche, B.; Zheng, M.; Choudhuri, A.; Chen, T.; and Wu, Z. 2025 b . 7DGS: Unified Spatial-Temporal-Angular Gaussian Splatting. In ICCV

    2025

  4. [12]

    N.; Chen, T.; and Wu, Z

    Gao, Z.; Planche, B.; Zheng, M.; Choudhuri, A.; Nguyen, V. N.; Chen, T.; and Wu, Z. 2026. Render- FM : Feedforward Model for Real-time Photorealistic Volumetric Rendering. In ECCV. ArXiv:2505.17338

    2026 arXiv

  5. [13]

    Hahlbohm, F.; Franke, L.; Eisemann, M.; and Magnor, M. 2026. Faster-GS : Analyzing and Improving Gaussian Splatting Optimization. In CVPR

    2026

  6. [14]

    Hanson, A.; Tu, A.; Lin, G.; Singla, V.; Zwicker, M.; and Goldstein, T. 2025. Speedy-Splat: Fast 3D Gaussian Splatting with Sparse Pixels and Sparse Primitives. In CVPR, 21537--21546

    2025

  7. [15]

    Huang, Z.; Wu, C.-Y.; Guo, Y.; Huang, X.; and Ren, L. 2026. 3DGEER : 3D G aussian Rendering Made Exact and Efficient for Generic Cameras. In ICLR

    2026

  8. [16]

    Jain, U.; Mirzaei, A.; and Gilitschenski, I. 2024. GaussianCut : Interactive Segmentation via Graph Cut for 3D G aussian Splatting. In NeurIPS

    2024

  9. [17]

    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): 139:1--139:14

    2023

  10. [18]

    Kleinbeck, C.; Schieber, H.; Engel, K.; Gutjahr, R.; and Roth, D. 2025. Multi-Layer G aussian Splatting for Immersive Anatomy Visualization. IEEE Transactions on Visualization and Computer Graphics, 31(5): 2353--2363

    2025

  11. [19]

    H.; and Botha, C

    Kroes, T.; Post, F. H.; and Botha, C. P. 2012. Exposure Render: An Interactive Photo-Realistic Volume Rendering Framework. PLoS ONE, 7(7): e38586

    2012

  12. [20]

    Y.-K.; Heng, P.-A.; and Dou, Q

    Li, C.; Tong, Y.; Chen, K.; Yang, Z.; Li, R.; Qiu, S.; Chan, J. Y.-K.; Heng, P.-A.; and Dou, Q. 2025 a . ClipGS : Clippable G aussian Splatting for Interactive Cinematic Visualization of Volumetric Medical Data. In MICCAI

    2025

  13. [21]

    Li, D.; Jia, D.; Rajeh, Y.; Engel, D.; and Viola, I. 2025 b . RaRa C lipper: A Clipper for G aussian Splatting Based on Ray Tracer and Rasterizer. In SIGGRAPH Asia 2025 Conference Papers

    2025

  14. [22]

    Li, H.; Liu, J.; Sznaier, M.; and Camps, O. 2025 c . 3D-HGS : 3D Half- G aussian Splatting. In CVPR, 10996--11005

    2025

  15. [23]

    Liang, Z.; Zhang, Q.; Hu, W.; Zhu, L.; Feng, Y.; and Jia, K. 2024. Analytic-Splatting: Anti-Aliased 3D G aussian Splatting via Analytic Integration. In ECCV

    2024

  16. [24]

    Liao, Z.; Ding, J.; Cui, S.; Gong, R.; Hu, B.; Wang, Y.; Li, H.; Wang, H.; Zhang, X.; and Fu, R. 2025. TC-GS : A Faster Gaussian Splatting Module Utilizing Tensor Cores. In SIGGRAPH Asia 2025 Conference Papers, 181:1--181:9

    2025

  17. [25]

    N.; Zheng, M.; Choudhuri, A.; Chen, T.; Wang, Y.; Feng, A.; and Wu, Z

    Liu, R.; Gao, Z.; Planche, B.; Chen, M.; Nguyen, V. N.; Zheng, M.; Choudhuri, A.; Chen, T.; Wang, Y.; Feng, A.; and Wu, Z. 2026. Universal Beta Splatting. In ICLR. ArXiv:2510.03312

    2026

  18. [26]

    Liu, R.; Sun, D.; Chen, M.; Wang, Y.; and Feng, A. 2025. Deformable beta splatting. In SIGGRAPH

    2025

  19. [27]

    Martschinke, J.; Hartnagel, S.; Keinert, B.; Engel, K.; and Stamminger, M. 2019. Adaptive Temporal Sampling for Volumetric Path Tracing of Medical Data. Computer Graphics Forum, 38(4): 67--76

    2019

  20. [28]

    Niedermayr, S.; Neuhauser, C.; Petkov, K.; Engel, K.; and Westermann, R. 2024. Application of 3D G aussian Splatting for Cinematic Anatomy on Consumer Class Devices. In Vision, Modeling, and Visualization (VMV). The Eurographics Association. ArXiv:2404.11285

    2024 arXiv

  21. [29]

    Nov \'a k, J.; Georgiev, I.; Hanika, J.; and Jarosz, W. 2018. Monte C arlo Methods for Volumetric Light Transport Simulation. Computer Graphics Forum, 37(2): 551--576. Proc. Eurographics State of the Art Reports

    2018

  22. [30]

    Peng, T.; Zha, R.; Li, Z.; Liu, X.; and Zou, Q. 2025. Three-Dimensional MRI Reconstruction with G aussian Representations: Tackling the Undersampling Problem. arXiv preprint arXiv:2502.06510

    2025

  23. [31]

    Tallis, G. M. 1961. The moment generating function of the truncated multi-normal distribution. Journal of the Royal Statistical Society: Series B (Methodological), 23(1): 223--229

    1961

  24. [32]

    C.; Sheikh, H

    Wang, Z.; Bovik, A. C.; Sheikh, H. R.; and Simoncelli, E. P. 2004. Image Quality Assessment: From Error Visibility to Structural Similarity. IEEE Transactions on Image Processing, 13(4): 600--612

    2004

  25. [33]

    Weiskopf, D.; Engel, K.; and Ertl, T. 2003. Interactive Clipping Techniques for Texture-Based Volume Visualization and Volume Shading. IEEE Transactions on Visualization and Computer Graphics, 9(3): 298--312

    2003

  26. [34]

    Xu, J.; Thevenon, G.; Chabat, T.; McCormick, M.; Li, F.; Birdsong, T.; Martin, K.; Lee, Y.; and Aylward, S. 2023. Interactive, in-browser cinematic volume rendering of medical images. Computational Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 11...

    2023

  27. [35]

    Yang, G.; Kie , S.; Luo, H.; Liu, X.; and Simon, S. 2026. Exact-GS : Mathematically Rigorous and Accurate 3D G aussian Splatting for 3D X -ray Reconstruction. In CVPR, 4902--4911

    2026

  28. [36]

    Yang, S.; Li, Q.; Shen, D.; Gong, B.; Dou, Q.; and Jin, Y. 2024. Deform3DGS : Flexible Deformation for Fast Surgical Scene Reconstruction with G aussian Splatting. In MICCAI

    2024

  29. [37]

    Ye, M.; Danelljan, M.; Yu, F.; and Ke, L. 2024. G aussian Grouping: Segment and Edit Anything in 3D Scenes. In ECCV

    2024

  30. [38]

    Yu, Z.; Chen, A.; Huang, B.; Sattler, T.; and Geiger, A. 2024. Mip-Splatting: Alias-free 3D Gaussian Splatting. In CVPR

    2024

  31. [39]

    J.; Cai, Y.; Cao, J.; Zhang, Y.; and Li, H

    Zha, R.; Lin, T. J.; Cai, Y.; Cao, J.; Zhang, Y.; and Li, H. 2024. R2-Gaussian : Rectifying Radiative G aussian Splatting for Tomographic Reconstruction. In NeurIPS

    2024

  32. [40]

    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 CVPR, 586--595

    2018

  33. [41]

    Zhou, S.; Chang, H.; Jiang, S.; Fan, Z.; Zhu, Z.; Xu, D.; Chari, P.; You, S.; Wang, Z.; and Kadambi, A. 2024. Feature 3DGS : Supercharging 3D G aussian Splatting to Enable Distilled Feature Fields. In CVPR

    2024

  34. [42]

    Zwicker, M.; Pfister, H.; Van Baar, J.; and Gross, M. 2001. EWA volume splatting. In Proceedings Visualization, 2001. VIS'01., 29--36. IEEE

    2001

Pith tools

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