Pith. sign in

REVIEW 5 major objections 4 minor 43 references

Gaussian Volumetric Representation for Efficient Shear-Warp Visualization

T0 review · 5 major / 4 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read A sparse set of learnable Gaussian kernels can approximate a dense medical volume well enough to render it in real time via shear-warp compositing at up to 43.86 frames per second, with compression ratios around 11:1.

desk verdict Reasonable combination of known pieces, but the unbiasedness claim rests on an unshown normalization and the tables don't add up. read the letter →

arxiv 2607.25377 v1 pith:FNJ7F5X5 submitted 2026-07-28 cs.CV cs.GR

Mayuri Mathur (1) , Ojaswa Sharma (1) ((1) Indraprastha Institute of Information Technology Delhi) This is my paper · ORCID
classification cs.CVcs.GR
keywords GaussianvolumetricrepresentationMonteCarloestimationcurriculumlearningshear-warprenderingmedicalvolumevisualizationcompressionMRIcryosection
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 a compact Gaussian mixture field — a sum of colored, anisotropic kernels — can stand in for a dense voxel grid in medical volume visualization. Trained with an importance-weighted Monte Carlo estimator, the representation is optimized on a small fraction of voxels while remaining, in expectation, consistent with the full-volume reconstruction objective. A curriculum schedule gradually introduces slice-plane supervision, giving the field the planar coherence that shear-warp rendering later exploits. The authors report that the resulting representation renders MRI and cryosection volumes at up to 43.86 frames per second with roughly 11:1 compression, while preserving anatomical structure; they also note that the smooth Gaussian kernels can blur sharp boundaries and occasionally produce small dot artifacts under heavy sparsity. If correct, the method offers a route to interactive exploration of large volumetric scans on ordinary GPUs.

What carries the argument

The central object is the Gaussian mixture field f(x) (Eq. 1), a sum of m colored anisotropic Gaussians with learnable means, covariances, colors, and opacities. The load-bearing mechanism is the importance-weighted Monte Carlo estimator (Eq. 4), which makes the sparse-voxel loss an unbiased proxy for the dense MSE loss; the curriculum schedule (Eq. 9) blends this voxel loss with a slice-loss term to inject planar coherence. The renderer is a shear-warp compositor that evaluates slice textures directly from the Gaussian field and reuses a single adaptive stack as long as the viewing direction stays within an angular threshold.

What would settle it

Train the proposed Gaussian representation on a synthetic volume with a known, analytically computable dense loss, and compare the average of the Monte Carlo loss (with the paper's importance weighting) against the true dense loss over many independent sample sets. If the estimator is unbiased, the two should match within sampling error; any systematic gap — for instance, because the actual sampling distribution differs from the normalized one used in the weights — reveals the bias and invalidates the claimed equivalence to dense training. A second check is to reproduce the reported PSNR value

Watch

Extended reading notes

Core claim

The central claim is that the continuous volumetric field f(x) = Σᵢ αᵢ Cᵢ exp(−½ (x−μᵢ)ᵀ Σᵢ⁻¹ (x−μᵢ)) defined by m learnable Gaussians can reconstruct a dense voxel grid from sparse supervision. The paper shows that the Monte Carlo estimator L̂_MC = (1/m) Σ (1/(n p(xᵢ))) ℓ(xᵢ; Θ) is an unbiased estimate of the dense MSE loss, so stochastic gradient descent on sparse voxel samples converges to the same optimum as dense training. It then augments this voxel supervision with slice-plane supervision through a sigmoid-gated curriculum, and renders the learned field with a shear-warp algorithm that composites slice textures sampled from the Gaussian field. On BraTS MRI, embryonic mouse MRI, and cr

Load-bearing premise

The sparse-training guarantee requires that the voxel-sampling distribution p(x) used to draw samples exactly matches the distribution used in the importance weights 1/(n p(x)), and that this match survives the periodic refresh of high-gradient voxels; the paper defers the construction of this distribution to a supplementary document that is not present.

Editorial extensions

If this is right

  • Medical volumes can be compressed by roughly an order of magnitude without visible loss of anatomical detail, making storage and transmission of large datasets cheaper.
  • Interactive visualization of full-resolution MRI and cryosection volumes becomes possible on a single GPU, enabling real-time slice browsing, rotation, and organ-focused views.
  • The unbiased Monte Carlo training scheme suggests that any dense voxel objective can be optimized from a sparse subset of voxels, as long as the sampling distribution is known and normalized.
  • The curriculum that moves from global voxel coverage to structured slice supervision provides a recipe for training implicit volumetric representations from sparse labels.
  • The adaptive single-stack shear-warp renderer reduces memory overhead compared to classic three-stack shear-warp, and could be applied to any analytic continuous field.

Reading between the lines

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

  • If the unbiasedness claim survives scrutiny, the same importance-weighting trick could be applied to time-resolved (4D) volumes by treating the temporal axis as an additional coordinate, with the curriculum progressively injecting temporal slice constraints.
  • The reported frame rates reflect rendering the Gaussian field; a fair end-to-end comparison with neural volume rendering would need to account for the cost of regenerating the slice stack whenever the camera direction changes, which the paper only bounds by an angular threshold.
  • Because importance sampling is driven by reconstruction error, the method may underperform on volumes with fine, low-gradient textures such as noisy soft tissue; a controlled experiment on synthetic volumes with known frequency content would map this failure mode.
  • The compression ratio counts the Gaussian parameter footprint against the original volume; a practical deployment that maintains slice stacks on demand may carry additional memory cost, so the reported ratio could be optimistic for a fully interactive system.
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 / 4 minor

Summary. The paper proposes a Gaussian mixture representation f(x) (Eq. 1) for dense medical volumes, trained with an importance-weighted Monte Carlo estimate of the dense mean-squared-error objective (Eqs. 2–6) combined with a curriculum that mixes sparse voxel supervision and slice-plane supervision (Eqs. 9–10). The learned field is rendered with a shear-warp GPU renderer. Experiments on BraTS MRI, mouse MRI, and Visible Korean cryosection volumes report high PSNR/SSIM/MS-SSIM, compression ratios of 4.73:1–11.31:1, and rendering speeds up to 43.86 FPS.

Significance. If the theoretical and empirical claims are correct, the paper is a useful step toward real-time, memory-efficient visualization of large medical volumes. The core Monte Carlo construction in Eq. (4) is standard and the unbiasedness identity (5) is correctly derived for a fixed, normalized sampling distribution. The curriculum idea — using sparse voxels for global coverage and slice samples for spatial coherence — is well-motivated by the shear-warp rendering target. The paper also ships a concrete rendering-speed claim and a systematic ablation. However, the manuscript as written does not establish the practical unbiasedness of the training objective because the sampling distribution is deferred to a missing supplement, and several quantitative claims are internally inconsistent.

major comments (5)
  1. [§3.2, Eqs. (7)–(8)] The estimator (4) is unbiased only if p is the exact normalized distribution from which the samples are drawn. Eq. (8) defines p_importance ∝ ℓ(x;Θ)+ε, whose normalization constant requires summing over all n voxels — i.e., a dense pass. The text states that details are in the supplementary material, but the supplement is not present. If Z is computed over a subset, or if samples are drawn from a stale/approximate distribution, Eq. (5) fails and the objective is biased. Please provide the exact normalization and sampling protocol, or use a self-normalized estimator and quantify the resulting bias.
  2. [§3.2, Eq. (6)] The gradient estimator is claimed to satisfy E[∇Θ L̂MC] = ∇Θ Ldense. However, p in Eqs. (7)–(8) depends on Θ through ℓ(x;Θ). When p is refreshed during training, the derivative ∂p/∂Θ is nonzero, and Eq. (6) omits the term -ℓ(x;Θ) ∇p(x;Θ)/p(x;Θ)^2 inside the expectation. Thus the expected gradient is ∇Ldense - (1/n)Σ_x ℓ(x;Θ)∇p(x;Θ)/p(x;Θ), which is not ∇Ldense unless p is fixed. The convergence claim is therefore not justified as stated. Please clarify whether p is treated as a fixed proposal or correct the estimator/analysis.
  3. [Table 2] The reported PSNR and MSE values are mutually inconsistent. For normalized intensities (peak = 1), PSNR = 10 log10(1/MSE). In the Ours row, reconstruction PSNR 34.81 dB with MSE 3.98e-4 gives 34.0 dB, and rendering PSNR 54.87 dB with MSE 2.1e-5 gives 46.8 dB. Similar inconsistencies appear in baseline rows (e.g., Voxel MLP: 23.63 dB vs 4.34e-4 gives 33.6 dB). If peak = 255 is used instead, all values are offset by 48.1 dB. Please specify the PSNR convention and recompute the table, and add error bars across multiple training runs.
  4. [§4, Table 1] The implementation states that the sparse voxel set is chosen so that the Gaussian memory footprint remains within 25% of the original volume size, implying at most a 4:1 compression ratio. Table 1 reports compression ratios of 4.73:1, 9.34:1, 9.45:1, 10.09:1, 11.31:1, and 9.94:1, all exceeding 4:1. Please clarify how the Gaussian parameter memory is counted (parameter count, storage precision, inclusion of the voxel indices, etc.) and reconcile this with the 25% budget statement.
  5. [§4.1, Related Work] The closest Gaussian-based medical volume methods — iVR-GS [33], Kleinbeck et al. [13], ClipGS [16], and Condor et al. [5] — are discussed in Related Work but are not included in Table 2. Because the central claim is that a Gaussian representation enables efficient real-time visualization, a quantitative comparison against these methods is needed to establish the advantage. Please either add these baselines or clearly explain why they are inapplicable to the sparse-supervision setting, with numbers if possible.
minor comments (4)
  1. [§3.3, Eq. (9)] The schedule λs(t) = λs,max σ((t-γs)/βs) is not evaluated at t=0 in the text; please specify the initial slice-loss weight and how λv(t) is computed in the transition region.
  2. [Table 2] The table header abbreviates MS-SSIM as MS-SIM in some columns, and FLOPs units (×10^9, ×10^12, etc.) are inconsistent in formatting. Please standardize and define all abbreviations.
  3. [§3.4] The angular threshold for slice-stack refresh is mentioned but never reported or ablated. Since it directly affects FPS vs. quality trade-off, please give the value used in the experiments.
  4. [§4] Please report exact volume dimensions, preprocessing (e.g., normalization to [0,1] or [0,255]), and the GPU model used for the timing experiments, as rendering FPS is hardware-dependent.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the Monte Carlo estimator is a standard unbiasedness argument and all reported performance numbers are empirical.

full rationale

The derivation chain is self-contained. Equation (4) defines an importance-weighted Monte Carlo estimator, and Eq. (5) computes its expectation under p(x); for any properly normalized proposal p(x) from which samples are drawn, E[L_hat_MC] = L_dense by construction. The fact that Eq. (8) chooses p_importance proportional to the reconstruction error ℓ(x;Θ) is valid importance sampling—it does not make the estimate equal to the fit, because the 1/p(x) weight cancels exactly in expectation. The reported PSNR, SSIM, FPS, and compression ratios (Tables 1–3) are measured on the reconstructed volumes/renders rather than derived from fitted constants, so there is no fitted-input-called-prediction step. No load-bearing self-citation appears: references to Gaussian splatting, Monte Carlo methods, and curriculum learning are external, and the authors do not invoke a prior uniqueness theorem or ansatz of their own. The main caveat is an omitted proof: Section 3.2 defers "The details of our sampling and the corresponding probability estimates" to a supplement that is not present, leaving the practical normalization of Eq. (8) unverified. That is a completeness/robustness gap, not a circular reduction, because the unbiasedness identity does not rely on the specific normalized p to hold; it relies only on p being the true sampling distribution. Evaluation against the same volumes used for supervision is standard for per-volume compression/representation tasks and does not make the empirical claims circular.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical or ontological entities; the Gaussian kernels are an existing primitive (3D Gaussian splatting) repurposed for volume fields, with no independent falsifiable handle outside the paper. The load-bearing modeling choices are the smooth Gaussian mixture, the gradient/random sparse sampling split, and the curriculum schedule, all listed above as axioms and free parameters.

free parameters (7)
  • Sparse memory budget (25% of original volume) = 25% (deferred)
    Sets the number of Gaussians / supervised voxels; directly determines compression ratios reported (4.73:1-11.31:1); no sensitivity analysis.
  • High-gradient/random split (60/40) = 60% gradient / 40% random
    Chosen by hand in Sec. 4; no ablation on this ratio; controls coverage of informative regions.
  • Importance-balance lambda in Eq (7) = not reported
    Balances uniform vs error-based sampling; value not given; unbiasedness claim depends on how p is normalized.
  • Curriculum schedule lambda_s,max, gamma_s, beta_s = not reported
    Sigmoid schedule constants in Eq (9); not reported; ablation shows schedule order matters, so these affect results.
  • Angular threshold for slice-stack refresh = not reported
    Shear-warp renderer reuses slice stack when camera rotation below threshold; threshold not specified; affects FPS and image fidelity as camera moves.
  • epsilon in p_importance (Eq (8)) = not reported
    Ensures non-zero probability for importance sampling; value not reported.
  • Gaussian count m / parameter count = not reported
    Number of Gaussian kernels initialized from the sparse voxel set not specified; central to compression ratio and quality.
assumptions (5)
  • standard math Monte Carlo estimator unbiasedness with known normalized sampling distribution p(x) (Eq 5)
    Textbook result, but its application assumes p(x_i) is exactly the probability mass used to draw samples at every iteration.
  • domain assumption A mixture of smooth Gaussian kernels with constant per-kernel color can approximate dense medical volumes to the reported fidelity within the memory budget
    Section 3.1, Eq (1). The entire method rests on expressive power of this representation; the paper notes blurring near sharp boundaries itself.
  • domain assumption Sparse voxels (60% high-gradient, 40% random) are sufficient to reconstruct anatomy
    Section 4. No proof that gradient-magnitude sampling covers all radiometric statistics; no error bars.
  • ad hoc to paper Pointwise L2 loss on sparse pixels from a plane imposes spatial coherence/geometric structure
    Section 3.3. The slice loss is identical pointwise loss; the claimed 'structured constraints' are not formalized; no spatial regularization term exists.
  • domain assumption Shear-warp compositing from a single adaptive slice stack within angular threshold equals full volume rendering
    Section 3.4. Stack reuse approximation; angular threshold unspecified; image error vs camera angle not measured.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Gaussian Volumetric Representation for Efficient Shear-Warp Visualization." pith.science (2026). https://pith.science/paper/FNJ7F5X5

@misc{pith2026260725377,
  author       = {Pith},
  title        = {Pith review of: Gaussian Volumetric Representation for Efficient Shear-Warp Visualization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FNJ7F5X5}},
  note         = {Machine review of arXiv:2607.25377}
}
read the original abstract

Medical image visualization requires volumetric rendering algorithms that preserve anatomical fidelity while maintaining high rendering speeds. To address the high computational cost of large volumetric datasets, we propose a Gaussian-based volumetric representation for efficient visualization of dense medical volumes without compromising structural and radiometric details. We optimize the proposed representation using Monte Carlo volumetric estimation, which enables training on a highly sparse subset of voxels while maintaining consistency with the dense volumetric objective. In addition, we introduce a curriculum learning strategy that progressively incorporates structured slice-based sampling during training. Sparse voxel samples provide an early global coverage of the volume, while slice samples capture spatially correlated regions that aid geometric structure and texture continuity. This combination enables the Gaussian representation to learn anatomical details of various structures and corresponding textures from sparse supervision while significantly reducing the computational cost associated with dense voxel processing. The learned representation supports slice-based rendering methods such as shear-warp volume rendering, enabling efficient visualization of multimodal medical datasets including MRI and Cryosection volumes while preserving anatomical structures. Using sparse supervision, our method achieves up to 43.86 FPS rendering with a compression ratio of 11.31:1.

Figures

Figures reproduced from arXiv: 2607.25377 by the authors.

Figure 1
Figure 1. Overview of the proposed Gaussian-based volumetric representation for real￾time rendering. Sparse voxel and slice samples are used under a curriculum sampling strategy to optimize a continuous Gaussian volumetric representation, which is sub￾sequently used to generate arbitrary view-based slice textures for shear–warp volume rendering. the dense loss Ldense is impractical for large volumetric datasets where grid siz… view at source ↗
Figure 2
Figure 2. Visual comparison of rendered views for different organs, including brain, liver, heart, and lungs from Cryosection data, BraTS MRI data, and mouse neonatal MRI data. Datasets The evaluation metrics are computed for grayscale MRI modali￾ties, including the BraTS [23] dataset and the embryonic–neonatal mouse MRI dataset. The RGB modality corresponds to the Cryosection dataset, along with organ-level volumes including… view at source ↗
Figure 3
Figure 3. Comparisons with learning-based volumetric representation methods [PITH_FULL_IMAGE:figures/full_fig_p013_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Visual comparisons of the ablation study on cross sections and rendered views from the shear-warp renderer [PITH_FULL_IMAGE:figures/full_fig_p013_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 1 canonical work pages

  1. [33]

    IEEE Transactions on Visualization and Computer Graphics (2025)

    Tang, K., Yao, S., Wang, C.: iVR-GS: Inverse volume rendering for explorable visu- alization via editable 3D Gaussian splatting. IEEE Transactions on Visualization and Computer Graphics (2025)

    2025

  2. [13]

    IEEE Transactions on Visualization and Computer Graphics (2025)

    Kleinbeck, C., Schieber, H., Engel, K., Gutjahr, R., Roth, D.: Multi-layer gaussian splatting for immersive anatomy visualization. IEEE Transactions on Visualization and Computer Graphics (2025)

    2025

  3. [16]

    arXiv preprint arXiv:2507.06647 (2025)

    Li, C., Tong, Y., Chen, K., Yang, Z., Li, R., Qiu, S., Chan, J.Y.K., Heng, P.A., Dou, Q.: Clipgs: Clippable gaussian splatting for interactive cinematic visualization of volumetric medical data. arXiv preprint arXiv:2507.06647 (2025)

    arXiv 2025

  4. [5]

    ACM Transactions on Graphics44(1), 1–17 (2025)

    Condor, J., Speierer, S., Bode, L., Bozic, A., Green, S., Didyk, P., Jarabo, A.: Don’t Splat your Gaussians: Volumetric Ray-Traced Primitives for Modeling and Rendering Scattering and Emissive Media. ACM Transactions on Graphics44(1), 1–17 (2025)

    2025

  5. [1]

    Mathur and O

    Ahmed,N.,Natarajan,T.,Rao,K.R.:Discretecosinetransform.IEEEtransactions on Computers100(1), 90–93 (1974) 16 M. Mathur and O. Sharma

    1974

  6. [2]

    ACM Transactions on Multimedia Computing, Communications and Applications (2026)

    Chang, Z., Jin, H., Song, Y., Sun, Y., Yu, H.: GAT-NeRF: Geometry-aware- transformer enhanced neural radiance fields for high-fidelity 4d facial avatars. ACM Transactions on Multimedia Computing, Communications and Applications (2026)

    2026

  7. [3]

    In: European Conference on Computer Vision

    Chen, Y., Wang, X.: Transformers as meta-learners for implicit neural representa- tions. In: European Conference on Computer Vision. pp. 170–187. Springer (2022)

    2022

  8. [4]

    In: International conference on medical image computing and computer-assisted intervention

    Çiçek, Ö., Abdulkadir, A., Lienkamp, S.S., Brox, T., Ronneberger, O.: 3D U-Net: learning dense volumetric segmentation from sparse annotation. In: International conference on medical image computing and computer-assisted intervention. pp. 424–432. Springer (2016)

    2016

Show all 43 references
  1. [6]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)

    Cong, W., Liang, H., Wang, P., Fan, Z., Chen, T., Varma, M., Wang, Y., Wang, Z.: Enhancing NeRF akin to Enhancing LLMs: Generalizable NeRF Transformer with Mixture-of-View-Experts. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). pp. 3193–3204 (2023)

    2023

  2. [7]

    In: 2022 44th annual international conference of the IEEE engineering in medicine & Biology society (EMBC)

    Corona-Figueroa,A.,Frawley,J.,Bond-Taylor,S.,Bethapudi,S.,Shum,H.P.,Will- cocks, C.G.: Mednerf: Medical neural radiance fields for reconstructing 3d-aware ct-projections from a single X-ray. In: 2022 44th annual international conference of the IEEE engineering in medicine & Bi...

    2022

  3. [8]

    In: International Symposium on Visual Computing

    Devkota,S.,Pattanaik,S.:Deeplearningbasedsuper-resolutionformedicalvolume visualization with direct volume rendering. In: International Symposium on Visual Computing. pp. 103–114. Springer (2022)

    2022

  4. [9]

    In: European Conference on Computer Vision

    Duan, Y., Zhu, H., Wang, H., Yi, L., Nevatia, R., Guibas, L.J.: Curriculum deepsdf. In: European Conference on Computer Vision. pp. 51–67. Springer (2020)

    2020

  5. [10]

    Kerbl, B., Kopanas, G., Leimkühler, T., Drettakis, G.: 3D Gaussian splatting for real-time radiance field rendering (2023),https://arxiv.org/abs/2308.04079

    2023 arXiv

  6. [11]

    International Journal of Computer Assisted Radiology and Surgery20(7), 1481–1490 (2025)

    Khojasteh, S.B., Fuentes-Jimenez, D., Pizarro, D., Espinel, Y., Bartoli, A.: MIS- NeRF: neural radiance fields in minimally-invasive surgery. International Journal of Computer Assisted Radiology and Surgery20(7), 1481–1490 (2025)

    2025

  7. [12]

    IEEE Access9, 124281–124294 (2021)

    Kim, S., Jang, Y., Kim, S.E.: Image-based TF colorization with CNN for direct volume rendering. IEEE Access9, 124281–124294 (2021)

    2021

  8. [14]

    SIAM review 51(3), 455–500 (2009)

    Kolda, T.G., Bader, B.W.: Tensor decompositions and applications. SIAM review 51(3), 455–500 (2009)

    2009

  9. [15]

    In: Proceedings of the 21st annual conference on Computer graphics and interactive techniques

    Lacroute, P., Levoy, M.: Fast volume rendering using a shear-warp factorization of the viewing transformation. In: Proceedings of the 21st annual conference on Computer graphics and interactive techniques. pp. 451–458 (1994)

    1994

  10. [17]

    Advances in neural information processing systems36, 62332–62351 (2023)

    Liang, Y., He, H., Chen, Y.: Retr: Modeling rendering via transformer for gener- alizable neural surface reconstruction. Advances in neural information processing systems36, 62332–62351 (2023)

    2023

  11. [18]

    In: European conference on computer vision

    Liang, Z., Zhang, Q., Hu, W., Zhu, L., Feng, Y., Jia, K.: Analytic-splatting: Anti- aliased 3D Gaussian splatting via analytic integration. In: European conference on computer vision. pp. 281–297. Springer (2024) Gaussian Volumetric Shear-Warp Visualization 17

    2024

  12. [19]

    Lin, K.E., Lin, Y.C., Lai, W.S., Lin, T.Y., Shih, Y.C., Ramamoorthi, R.: Vision transformerfornerf-basedviewsynthesisfromasingleinputimage.In:Proceedings of the IEEE/CVF winter conference on applications of computer vision. pp. 806– 815 (2023)

    2023

  13. [20]

    EURASIP Journal on Advances in Signal Processing2020(1), 25 (2020)

    Luengo, D., Martino, L., Bugallo, M., Elvira, V., Särkkä, S.: A survey of Monte Carlo methods for parameter estimation. EURASIP Journal on Advances in Signal Processing2020(1), 25 (2020)

    2020

  14. [21]

    Lyu, J., Ling, S.H., Banerjee, S., Zheng, J., Lai, K.L., Yang, D., Zheng, Y.P., Bi, X., Su, S., Chamoli, U.: Ultrasound volume projection image quality selection by rankingfromconvolutionalRankNet.ComputerizedMedicalImagingandGraphics 89, 101847 (2021)

    2021

  15. [22]

    Computer graphics and image processing19(2), 129–147 (1982)

    Meagher, D.: Geometric modeling using octree encoding. Computer graphics and image processing19(2), 129–147 (1982)

    1982

  16. [23]

    IEEE Transactions on Medical Imaging34(10), 1993–2024 (2015).https://doi.org/10.1109/TMI.2014.2377694

    Menze, B.H., Jakab, A., Bauer, S., Kalpathy-Cramer, J., Farahani, K., Kirby, J., Burren, Y., Porz, N., Slotboom, J., Wiest, R., Lanczi, L., Gerstner, E., Weber, M.A., Arbel, T., Avants, B.B., Ayache, N., Buendia, P., Collins, D.L., Cordier, N., Corso, J.J., Criminisi, A., Das,...

    1993

  17. [24]

    Commu- nications of the ACM65(1), 99–106 (2021)

    Mildenhall, B., Srinivasan, P.P., Tancik, M., Barron, J.T., Ramamoorthi, R., Ng, R.: Nerf: Representing scenes as neural radiance fields for view synthesis. Commu- nications of the ACM65(1), 99–106 (2021)

    2021

  18. [25]

    ACM65(1), 99–106 (Dec 2021).https://doi.org/10.1145/3503250,https:// doi.org/10.1145/3503250

    Mildenhall, B., Srinivasan, P.P., Tancik, M., Barron, J.T., Ramamoorthi, R., Ng, R.:NeRF:representingscenesasneuralradiancefieldsforviewsynthesis.Commun. ACM65(1), 99–106 (Dec 2021).https://doi.org/10.1145/3503250,https:// doi.org/10.1145/3503250

    2021 doi

  19. [26]

    ACM Trans

    Müller, T., Mcwilliams, B., Rousselle, F., Gross, M., Novák, J.: Neural importance sampling. ACM Trans. Graph.38(5) (Oct 2019).https://doi.org/10.1145/ 3341156,https://doi.org/10.1145/3341156

    2019 doi

  20. [27]

    IEEE trans- actions on medical imaging24(3), 352–360 (2005)

    Park, J.S., Chung, M.S., Hwang, S.B., Lee, Y.S., Har, D.H., Park, H.S.: Visible Korean human: improved serially sectioned images of the entire body. IEEE trans- actions on medical imaging24(3), 352–360 (2005)

    2005

  21. [28]

    Proceedings of the National Academy of Sciences105(34), 12331–12336 (2008).https://doi.org/10.1073/ pnas.0805747105,https://www.pnas.org/doi/abs/10.1073/pnas.0805747105

    Petiet, A.E., Kaufman, M.H., Goddeeris, M.M., Brandenburg, J., Elmore, S.A., Johnson, G.A.: High-resolution magnetic resonance histology of the embryonic and neonatal mouse: A 4D atlas and morphologic database. Proceedings of the National Academy of Sciences105(34), 12331–1233...

    2008 doi

  22. [29]

    In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F

    Ronneberger, O., Fischer, P., Brox, T.: U-net: Convolutional networks for biomed- ical image segmentation. In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.) Medical Image Computing and Computer-Assisted Intervention – MICCAI

  23. [30]

    ACM Transactions on Graphics (TOG)41(4), 1–13 (2022) 18 M

    Rückert, D., Wang, Y., Li, R., Idoughi, R., Heidrich, W.: Neat: Neural adaptive tomography. ACM Transactions on Graphics (TOG)41(4), 1–13 (2022) 18 M. Mathur and O. Sharma

    2022

  24. [31]

    Journal of Computer and Commu- nications7(3), 8–18 (2019)

    Sara, U., Akter, M., Uddin, M.S., et al.: Image quality assessment through FSIM, SSIM, MSE and PSNR—a comparative study. Journal of Computer and Commu- nications7(3), 8–18 (2019)

    2019

  25. [32]

    In: ICASSP 2024-2024 IEEE In- ternational Conference on Acoustics, Speech and Signal Processing (ICASSP)

    Su, Y., Wang, S., Wang, H.: Dt-nerf: Decomposed triplane-hash neural radiance fields for high-fidelity talking portrait synthesis. In: ICASSP 2024-2024 IEEE In- ternational Conference on Acoustics, Speech and Signal Processing (ICASSP). pp. 3975–3979. IEEE (2024)

    2024

  26. [34]

    In: Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition

    Tang, Z.J., Cham, T.J., Zhao, H.: Able-nerf: Attention-based rendering with learn- able embeddings for neural radiance field. In: Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition. pp. 16559–16568 (2023)

    2023

  27. [35]

    Advances in neural information pro- cessing systems30(2017)

    Vaswani,A.,Shazeer,N.,Parmar,N.,Uszkoreit,J.,Jones,L.,Gomez,A.N.,Kaiser, Ł., Polosukhin, I.: Attention is all you need. Advances in neural information pro- cessing systems30(2017)

    2017

  28. [36]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)

    Wang, D., Cui, X., Chen, X., Zou, Z., Shi, T., Salcudean, S., Wang, Z.J., Ward, R.: Multi-View 3D Reconstruction With Transformers. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). pp. 5722–5731 (2021)

    2021

  29. [37]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Wang, Y., Gan, W., Yang, J., Wu, W., Yan, J.: Dynamic curriculum learning for imbalanced data classification. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 5017–5026 (2019)

    2019

  30. [38]

    In: Leibe, B., Matas, J., Sebe, N., Welling, M

    Wen, Y., Zhang, K., Li, Z., Qiao, Y.: A discriminative feature learning approach for deep face recognition. In: Leibe, B., Matas, J., Sebe, N., Welling, M. (eds.) Computer Vision – ECCV 2016. pp. 499–515. Springer International Publishing, Cham (2016)

    2016

  31. [39]

    In: Proceedings of the IEEE/CVF International Confer- ence on Computer Vision (ICCV) (2025)

    Xu, J., Gao, S., Shan, Y.: FreeSplatter: Pose-Free Gaussian Splatting for Sparse- View 3D Reconstruction. In: Proceedings of the IEEE/CVF International Confer- ence on Computer Vision (ICCV) (2025)

    2025

  32. [40]

    In: International Conference on Medical Image Computing and Computer-Assisted Intervention

    Zha, R., Zhang, Y., Li, H.: NAF: neural attenuation fields for sparse-view cbct reconstruction. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 442–452. Springer (2022)

    2022

  33. [41]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2024)

    Zhong, Y., Hong, L., Li, Z., Xu, D.: CVT-xRF: Contrastive In-Voxel Transformer for 3D Consistent Radiance Fields from Sparse Inputs. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2024)

    2024

  34. [42]

    In: Proceedings of the IEEE interna- tional conference on computer vision

    Zhu, J.Y., Park, T., Isola, P., Efros, A.A.: Unpaired image-to-image translation using cycle-consistent adversarial networks. In: Proceedings of the IEEE interna- tional conference on computer vision. pp. 2223–2232 (2017)

    2017

  35. [2015]

    pp. 234–241. Springer International Publishing, Cham (2015)

    2015

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.