Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

PS-GS: Gaussian Splatting for Multi-View Photometric Stereo

T0 review · 3 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read PS-GS is the first Gaussian-splatting method for multi-view photometric stereo, jointly estimating geometry, materials, and lighting from multi-light images in about 0.7 hours while achieving state-of-the-art normal accuracy and enabling…

desk verdict First 2DGS+MVPS integration with strong benchmark numbers; the UPS-normal regularizer is load-bearing and its GBR ambiguity is unaddressed, so the accuracy claims are conditional. read the letter →

arxiv 2507.18231 v1 pith:VZHUO5R5 submitted 2025-07-24 cs.GR cs.CV

Yixiao Chen , Bin Liang , Hanzhi Guo , Yongqing Cheng , Jiayi Zhao , Dongdong Weng This is my paper · ORCID
classification cs.GRcs.CV
keywords multi-viewphotometricstereoGaussiansplattinginverserenderingsurfacenormalestimationrelightingmaterialediting2Dray-tracinguncalibrated
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 introduces PS-GS, an inverse-rendering method that combines 2D Gaussian splatting with multi-view photometric stereo to reconstruct an object's geometry, surface materials, and incident lighting from multi-view images captured under multiple directional lights. It claims to be the first Gaussian-splatting-based MVPS inverse-rendering approach, and reports state-of-the-art normal accuracy (for example, MAE 3.68 on the Bear object versus 4.86 for the DPIR baseline), top or near-top novel-view synthesis and relighting quality, and far lower cost than prior MVPS renderers: about 0.7 hours and 6.5 GB of GPU memory, versus over 22 hours and 9.6 GB for PS-NeRF. The practical payoff is that accurate, editable 3D models with fine surface detail can be produced quickly on modest hardware, and the method degrades gracefully with as few as five training views. The paper's route to that result is to regularize splat-rendered normals with uncalibrated photometric stereo normals and to regularize a neural lighting field with visibility computed by ray-tracing the Gaussian model under a single directional light.

What carries the argument

The load-bearing machinery is the 2D Gaussian primitive, whose surface normal is defined directly as the cross product of its two tangent vectors, giving per-primitive geometry that participates in alpha-blended feature maps; the full rendering equation with a simplified Disney BRDF (diffuse albedo, metallic, roughness) evaluated via deferred shading; an MLP that predicts incident radiance as a function of position, tangents, normal, and in/out directions with Fourier-encoded position; and two regularizers. The first regularizer is the UPS-estimated normal map, aligned to world coordinates by a fixed transform and compared with the rendered normal map at the pixel level. The second is visibility computed by a 2D Gaussian ray-tracing technique modified to sample a single directional light, used to constrain the MLP's predicted lighting. Together these components let multi-view and multi-light observations constrain material, shape, and lighting simultaneously.

What would settle it

A reader could run PS-GS on a synthetic object with ground-truth normals and deliberately feed the UPS guidance normal maps through a fixed, non-identity 3x3 linear transform (the generalized bas-relief ambiguity) before the alignment step. If the final reconstructed normals and mesh shift predictably with that transform, the assumption is load-bearing; if the reconstruction remains accurate, the optimization is absorbing the ambiguity elsewhere.

Watch

Extended reading notes

Core claim

PS-GS's central claim is that 2D Gaussian splatting can carry the full inverse-rendering workload for multi-view photometric stereo without sacrificing the efficiency of splatting or the physical fidelity of a full rendering equation. After a standard 2DGS pretraining stage that fixes a reliable initial geometry, the method jointly optimizes per-Gaussian BRDF parameters (albedo, roughness, metallic), the Gaussian normals, and an MLP that predicts incident lighting, using deferred shading in image space for speed. Two regularizers anchor the otherwise ill-posed optimization: the rendered normal maps are pushed toward normals estimated by an uncalibrated photometric stereo method, and the MLP-predicted lighting is pushed toward visibility computed by a modified 2D Gaussian ray-tracer adapted to a single directional light. On the DiLiGenT-MV and PS-NeRF synthetic datasets, the method reports the best normal MAE on five of seven objects, the best PSNR on four, and the best SSIM on all seven, while completing training in 0.7 hours and 6.5 GB of GPU memory. The result, if accepted, means a single optimized model supports novel-view synthesis, relighting, and direct material and shape editing.

Load-bearing premise

The strongest load-bearing assumption is that the uncalibrated photometric stereo normals are recovered in true metric camera coordinates, so that the fixed camera-to-world transform in the normal regularization aligns them with the Gaussian model's world-space normals; if those normals carry an unknown linear ambiguity, that bias is injected directly into the optimized geometry.

Editorial extensions

If this is right

  • Training time of about 0.7 hours and GPU memory of 6.5 GB make MVPS-based inverse rendering practical on a single modest GPU, a sharp drop from PS-NeRF's more than 22 hours.
  • The same optimized model supports novel-view synthesis, relighting, and material and shape editing by directly editing Gaussian primitive attributes.
  • Quality degrades gracefully with sparse views: the method still produces usable reconstructions with five training views.
  • Ablations show both regularizers matter: dropping the UPS normal regularization smooths away detail, and dropping the lighting regularization bakes lighting into albedo.
  • Relying on multi-light inputs rather than fixed environment illumination yields sharper shadows and more detailed normals than prior splatting-based inverse renderers.

Reading between the lines

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

  • If the UPS normals are affected by the generalized bas-relief ambiguity, the fixed alignment transform could systematically skew the recovered geometry; a testable extension is to estimate that linear transform jointly instead of hard-coding it.
  • The two-stage design suggests a general recipe: pretrain a geometric splat model, then run deferred inverse rendering with external normal and visibility regularizers, which could transfer to other light configurations beyond single directional lights.
  • Because ray-tracing is performed only against the Stage-I geometry, large geometry changes in Stage II could make the visibility regularizer stale; re-running ray-tracing periodically would test whether that matters.
  • Scene-level effects such as inter-reflections between separate objects are out of scope; extending the method to multi-object scenes would require coupling the lighting MLP with light transport beyond the current single-directional-light visibility model.
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 / 4 minor

Summary. The paper presents PS-GS, a two-stage inverse rendering method based on 2D Gaussian Splatting for multi-view photometric stereo. Stage I pretrains a standard 2DGS model on light-averaged images; Stage II jointly optimizes geometry, BRDF materials, and incident lighting using the rendering equation, an MLP-based lighting model, a normal regularization from uncalibrated photometric stereo (UPS) estimates, and a visibility-based lighting regularizer from 2D Gaussian ray-tracing adapted to single directional lights. The authors report improved normal accuracy, NVS/relighting quality, and lower training time and GPU memory relative to PS-NeRF, DPIR, R3DG, and IRGS on DiLiGenT-MV and PS-NeRF synthesis datasets.

Significance. If the results are reproducible, this is a useful and timely contribution: it appears to be the first Gaussian-splatting-based MVPS inverse rendering method, and the efficiency gains (0.7 hours, 6.5 GB GPU memory) over PS-NeRF are substantial. The explicit 2DGS representation also enables material and geometry editing. However, the headline claims are conditional on the UPS-estimated normals being metrically correct in camera coordinates, an assumption the paper neither justifies nor tests; the main quantitative claims therefore need additional evidence before they can be considered established.

major comments (3)
  1. [Stage I: 2D Gaussian Pretraining, Eq. (10); Stage II loss, Eq. (15)] The normal regularization Ln,r transforms the UPS normals Ne only by the rigid camera-to-world transform Tc2w and then penalizes the rendered normals against that target. This is valid only if Ne is expressed in the same metric camera coordinates as the posed images. Uncalibrated photometric stereo recovers normals only up to a generalized bas-relief (GBR) ambiguity, i.e., an unknown 3x3 linear transform per view; the cited Chen et al. 2019 deep UPS method may select one member of the GBR family, but the paper neither shows that this member is the metric one nor cites a GBR-resolution step. This assumption is load-bearing: Table 3 attributes a large part of the final normal accuracy to Ln,r (MAE 7.02 without it versus 5.57 with it), and Table 2 reports state-of-the-art MAE values against ground-truth normals. If the UPS normals are GBR-warped, the reported MAE partly measures the quality of the UPS prior rather than the contribution of PS-GS. The authors should either add a GBR-resolution/canonicalization step, or provide a concrete test on the actual benchmark objects (e.g., report the MAE of the raw UPS normals against ground truth after only a rigid alignment on the DiLiGenT-MV test views) to demonstrate that the assumption holds.
  2. [Experimental] The paper repeatedly defers implementation details to a supplementary document and ships no code, but the empirical claims depend on several unspecified choices: the loss weights lambda_n,c, lambda_n,r, lambda_o, lambda_inc, the MLP architecture and Fourier encoding for the incident lighting network, how the light direction omega_in is obtained for each multi-light image, the optimizer and learning-rate schedule, and the number of training iterations. Without these details, the quantitative results in Tables 1-3 cannot be reproduced or independently checked, and the sensitivity of the central regularizers to their weights cannot be assessed. The authors should include these details in the main paper or make the supplementary material and code available.
  3. [Tables 1 and 2] All quantitative results are reported from single runs with no error bars or variance across seeds. Some comparisons are close (e.g., Bear PSNR 40.22 for DPIR versus 39.95 for PS-GS in Table 1, and Cow LPIPS 0.642 for DPIR versus 0.931 for PS-GS), so the state-of-the-art claims are not robustly supported without some measure of variability. The authors should report at least mean and standard deviation over multiple runs, or state if only one run was performed and temper the superiority claims accordingly.
minor comments (4)
  1. [Eq. (10)] The loss term Ln,r is written as a sum of vectors without an explicit norm; it should specify L1 or L2 loss, otherwise the loss value is not well defined.
  2. [Captions and text] There are several typos and caption errors: 'devided' and 'rednering' in the Preliminary section, 'uesd' in the rendering equation text, and Table 2 is captioned 'Qualitative comparison' even though it reports quantitative MAE values. These should be corrected.
  3. [Table 2] The rightmost columns 'GPU memory' and 'Time' are not normal-accuracy metrics and should be moved to a separate efficiency table or clearly labeled in the caption to avoid mixing metric types.
  4. [Eq. (14) and contribution claim] The paper describes the rendering model as the 'full rendering equation without simplification,' but Eq. (14) accounts only for direct illumination from a single directional light; there is no interreflection or global illumination term. The authors should soften this claim or clarify that 'full' refers only to the local BRDF evaluation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: PS-GS uses UPS normals as external supervision; its normal-accuracy claims are evaluated on held-out views and are not construction-equivalent to the regularizer.

full rationale

The paper's central pipeline is an optimization that fits 2D Gaussians, BRDF parameters, and an incident-light MLP to multi-view multi-light images. The only external prior is the UPS normal map, which is computed from the same training images but is not a quantity the paper claims to predict; Ln,r is a regularizer, not a fitted parameter later reported as a prediction. The headline normal MAE is computed on held-out novel views not used in training, so it tests generalization of the learned geometry rather than reproducing the UPS input. The incident-light regularization uses visibility from 2DGS ray-tracing and is an independent physical constraint. The GBR ambiguity of uncalibrated PS is an assumption about the external normal estimator, not a self-referential derivation step; it affects the validity of the supervision, not the circularity of the method. No load-bearing step reduces by construction to a fitted parameter or to a self-citation chain.

Assumptions & free parameters 3 free parameters · 5 assumptions · 1 invented entities

The central claim depends on the geometric and BRDF model assumptions of 2DGS and Disney, the correctness of UPS normals as metric ground truth, and the stability of geometry between stages. These are standard domain assumptions for the subfield, but the UPS ambiguity is the least justified and could systematically bias the results.

free parameters (3)
  • loss weights λn,c, λn,r, λo, λinc = not reported
    The total losses in Eqs. (9) and (15) depend on these hand-chosen weights; their values are not given in the main text, so the balance between photometric, normal, mask, and lighting terms is indeterminate.
  • incident lighting MLP weights θ = learned from data
    The lighting network fθ in Eq. (13) is a learned function of position, tangent, scale, normal, and light directions; its parameters are fit to the training images and are central to the lighting estimate.
  • Gaussian primitive BRDF parameters (albedo, metallic, roughness) = per-primitive, learned
    Each Gaussian carries learnable albedo, metallic, and roughness used in the Disney BRDF (Eq. 7); these are fitted to the multi-light images and are the material estimate the paper claims to recover.
assumptions (5)
  • domain assumption The object is illuminated exclusively by directional lights, with one light per image, and the BRDF is an isotropic Disney model with only albedo, metallic, and roughness.
    Used throughout Stage II (Eqs. 6-7, 14). Real objects may have non-isotropic reflectance, inter-reflections, or ambient light, which the model does not capture.
  • domain assumption The camera poses and object masks are known and correct.
    Stage I uses posed multi-view images and masks (Eq. 11). Inaccurate poses or masks degrade the initialized geometry and the subsequent UPS alignment.
  • ad hoc to paper UPS-estimated normals are metric-correct up to the rigid transform Tc2w; the generalized bas-relief ambiguity is absent.
    Eq. (10) and the Stage II normal regularization use UPS normals as guidance without resolving the GBR ambiguity inherent to uncalibrated photometric stereo. This is an unstated assumption unique to the paper's setup.
  • domain assumption The geometry changes only slightly in Stage II, so visibility computed by ray tracing on the Stage I pretrained model remains valid.
    The lighting regularization Linc is computed on the Stage I model: 'ray-tracing is only applied on the pretrained model obtained in Stage I, as the object's geometry exhibits slight changes in Stage II.' If geometry changes substantially, the visibility is stale.
  • domain assumption The 2D Gaussian disk representation can faithfully represent the object surface, including sharp edges and discontinuities.
    The method inherits 2DGS assumptions that smooth disks can approximate the surface; objects with complex topology or transparency are outside the stated scope ('object-level IR').
invented entities (1)
  • Implicit incident lighting MLP fθ
    purpose: Predicts incident radiance for each Gaussian as a function of position, tangents, scale, normal, and light directions, replacing explicit per-light/per-Gaussian lighting parameters.
    A learned model component rather than a physical entity; it has no falsifiable prediction outside the method's own reconstruction. The paper does not provide external evidence for its correctness beyond visual and quantitative fit on benchmarks.

how reviews work

0 comments
Cite this review

Pith. "Pith review of PS-GS: Gaussian Splatting for Multi-View Photometric Stereo." pith.science (2026). https://pith.science/paper/VZHUO5R5

@misc{pith2026250718231,
  author       = {Pith},
  title        = {Pith review of: PS-GS: Gaussian Splatting for Multi-View Photometric Stereo},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VZHUO5R5}},
  note         = {Machine review of arXiv:2507.18231}
}
read the original abstract

Integrating inverse rendering with multi-view photometric stereo (MVPS) yields more accurate 3D reconstructions than the inverse rendering approaches that rely on fixed environment illumination. However, efficient inverse rendering with MVPS remains challenging. To fill this gap, we introduce the Gaussian Splatting for Multi-view Photometric Stereo (PS-GS), which efficiently and jointly estimates the geometry, materials, and lighting of the object that is illuminated by diverse directional lights (multi-light). Our method first reconstructs a standard 2D Gaussian splatting model as the initial geometry. Based on the initialization model, it then proceeds with the deferred inverse rendering by the full rendering equation containing a lighting-computing multi-layer perceptron. During the whole optimization, we regularize the rendered normal maps by the uncalibrated photometric stereo estimated normals. We also propose the 2D Gaussian ray-tracing for single directional light to refine the incident lighting. The regularizations and the use of multi-view and multi-light images mitigate the ill-posed problem of inverse rendering. After optimization, the reconstructed object can be used for novel-view synthesis, relighting, and material and shape editing. Experiments on both synthetic and real datasets demonstrate that our method outperforms prior works in terms of reconstruction accuracy and computational efficiency.

Figures

Figures reproduced from arXiv: 2507.18231 by the authors.

Figure 1
Figure 1. Overview of PS-GS. Starting with a set of multi-view and multi-light images, we pretrain a standard 2D Gaussian model for geometry initialization. Based on the pretrained model, we jointly optimize the geometry, materials, and lighting. To encourage accurate surface reconstruction, we compute guidance normals by the uncalibrated photometric stereo (UPS) method to regularize the rendered normal maps from the Gaussian… view at source ↗
Figure 2
Figure 2. Qualitative comparison of novel view synthesis [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 5
Figure 5. Ablation study on normal regularization. The red [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figures from the paper (3 more)
Figure 6
Figure 6. Figure 6: Ablation study on incident lighting regularization. [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 8
Figure 8. Figure 8: Applications on PS-GS, including shape removal and material editing. For material editing, the first column is the result of replacing the diffuse map with the alternative, and the second column is the result of reducing the intensity of the specular map. Conclusion In…
Figure 7
Figure 7. Figure 7: Ablation study on training views. Applications Owing to the explicit Gaussian primitive representation, ma￾terials and geometry editing can be achieved conveniently through PS-GS by editing the parameters of Gaussian prim￾itives. Specifically, PS-GS allows material edi…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. A Hybrid Neural-Microfacet BRDF Model for Real-Time Rendering

    cs.GR 2026-08 conditional novelty 6.0 of 10

    A hybrid BRDF model, combining a GGX analytical term with a tiny learned residual and gating network, fits measured materials more accurately than fully neural models at equal memory cost.

Reference graph

Works this paper leans on

40 extracted references · 37 canonical work pages · cited by 1 Pith paper

  1. [1]

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

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    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 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    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 Proceedings of the IEEE/CVF International Conference on Computer Vision, 5835--5844. IEEE Computer Society

    work page 2021

  4. [4]

    B.; Durou, J.-D.; and Calvet, L

    Brument, B.; Bruneau, R.; Quéau, Y.; Mélou, J.; Lauze, F. B.; Durou, J.-D.; and Calvet, L. 2024. RNb-NeuS: Reflectance and Normal-Based Multi-View 3D Reconstruction. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 5230--5239. IEEE Computer Society

    work page 2024

  5. [5]

    Burley, B.; and Studios, W. D. A. 2012. Physically-based shading at Disney. In ACM SIGGRAPH

    work page 2012

  6. [6]

    Chen, G.; Han, K.; Shi, B.; Matsushita, Y.; and Wong, K.-Y. K. 2019. Self-Calibrating Deep Photometric Stereo Networks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE Computer Society

    work page 2019

  7. [7]

    Chung, H.-G.; Choi, S.; and Baek, S.-H. 2024. Differentiable Point-based Inverse Rendering. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 4399--4409. IEEE Computer Society

    work page 2024

  8. [8]

    Gao, J.; Gu, C.; Lin, Y.; Li, Z.; Zhu, H.; Cao, X.; Zhang, L.; and Yao, Y. 2024. Relightable 3D Gaussians: Realistic Point Cloud Relighting with BRDF Decomposition and Ray Tracing. In European conference on computer vision, 73--89. Springer

    work page 2024

Show all 40 references
  1. [9]

    Gu, C.; Wei, X.; Zeng, Z.; Yao, Y.; and Zhang, L. 2025. IRGS: Inter-Reflective Gaussian Splatting with 2D Gaussian Ray Tracing. In Proceedings of the Computer Vision and Pattern Recognition Conference (CVPR), 10943--10952. IEEE Computer Society

    2025

  2. [10]

    Hargreaves, S.; and Harris, M. 2004. Deferred shading. In Game Developers Conference

    2004

  3. [11]

    Hu, H.; Fan, Z.; Wu, T.; Xi, Y.; Lee, S.; Pavlakos, G.; and Wang, Z. 2025. Expressive Gaussian human avatars from monocular RGB video. In Proceedings of the 38th International Conference on Neural Information Processing Systems, NIPS '24. Red Hook, NY, USA: Curran Associates I...

    2025

  4. [12]

    Hu, T.; Liu, S.; Chen, Y.; Shen, T.; and Jia, J. 2022. EfficientNeRF Efficient Neural Radiance Fields. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 12902--12911. IEEE Computer Society

    2022

  5. [13]

    Hu, W.; Wang, Y.; Ma, L.; Yang, B.; Gao, L.; Liu, X.; and Ma, Y. 2023. Tri-MipRF: Tri-Mip Representation for Efficient Anti-Aliasing Neural Radiance Fields. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 19717--19726. IEEE Computer Society

    2023

  6. [14]

    Huang, B.; Yu, Z.; Chen, A.; Geiger, A.; and Gao, S. 2024. 2D Gaussian Splatting for Geometrically Accurate Radiance Fields. In ACM SIGGRAPH 2024 Conference Papers, 1--11

    2024

  7. [15]

    Ikehata, S. 2023. Scalable, Detailed and Mask-Free Universal Photometric Stereo. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 13198--13207. IEEE Computer Society

    2023

  8. [16]

    Jiang, Y.; Tu, J.; Liu, Y.; Gao, X.; Long, X.; Wang, W.; and Ma, Y. 2024. GaussianShader: 3D Gaussian Splatting with Shading Functions for Reflective Surfaces. In 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 5322--5332. IEEE Computer Society

    2024

  9. [17]

    Kaya, B.; Kumar, S.; Sarno, F.; Ferrari, V.; and Van Gool, L. 2022. Neural Radiance Fields Approach to Deep Multi-View Photometric Stereo. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 1965--1977. IEEE Computer Society

    2022

  10. [18]

    Kerbl, B.; Kopanas, G.; Leimkuehler, T.; and Drettakis, G. 2023. 3D Gaussian Splatting for Real-Time Radiance Field Rendering. ACM Trans. Graph., 42(4)

    2023

  11. [19]

    Li, M.; Zhou, Z.; Wu, Z.; Shi, B.; Diao, C.; and Tan, P. 2020. Multi-View Photometric Stereo: A Robust Solution and Benchmark Dataset for Spatially Varying Isotropic Materials. IEEE Transactions on Image Processing, 29: 4159--4173

    2020

  12. [20]

    Li, Z.; Chen, Z.; Li, Z.; and Xu, Y. 2024 a . Spacetime Gaussian Feature Splatting for Real-Time Dynamic View Synthesis. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 8508--8520. IEEE Computer Society

    2024

  13. [21]

    Li, Z.; Zheng, Z.; Wang, L.; and Liu, Y. 2024 b . Animatable Gaussians: Learning Pose-Dependent Gaussian Maps for High-Fidelity Human Avatar Modeling. In Processings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 19711--19722. IEEE Computer Society

    2024

  14. [22]

    Liang, Z.; Zhang, Q.; Feng, Y.; Shan, Y.; and Jia, K. 2024. GS-IR: 3D Gaussian Splatting for Inverse Rendering. In 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 21644--21653. IEEE Computer Society

    2024

  15. [23]

    Logothetis, F.; Mecca, R.; and Cipolla, R. 2019. A Differential Volumetric Approach to Multi-View Photometric Stereo. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 1052--1061. IEEE Computer Society

    2019

  16. [24]

    Matsuki, H.; Murai, R.; Kelly, P. H. J.; and Davison, A. J. 2024. Gaussian Splatting SLAM. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 18039--18048. IEEE Computer Society

    2024

  17. [25]

    P.; Tancik, M.; Barron, J

    Mildenhall, B.; Srinivasan, P. P.; Tancik, M.; Barron, J. T.; Ramamoorthi, R.; and Ng, R. 2020. NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis. In European Conference on Computer Vision, 405--421. Springer

    2020

  18. [26]

    M\" u ller, T.; Evans, A.; Schied, C.; and Keller, A. 2022. Instant neural graphics primitives with a multiresolution hash encoding. ACM Trans. Graph., 41(4)

    2022

  19. [27]

    N.; Matsushita, Y.; Tai, Y.-W.; and Kweon, I

    Park, J.; Sinha, S. N.; Matsushita, Y.; Tai, Y.-W.; and Kweon, I. S. 2017. Robust Multiview Photometric Stereo Using Planar Mesh Parameterization. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(8): 1591--1604

    2017

  20. [28]

    Shi, R.; Wei, X.; Wang, C.; and Su, H. 2024. ZeroRF: Fast Sparse View 360° Reconstruction with Zero Pretraining. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 21114--21124. IEEE Computer Society

    2024

  21. [29]

    C.; and Liu, Z

    Wang, G.; Chen, Z.; Loy, C. C.; and Liu, Z. 2023. SparseNeRF: Distilling Depth Ranking for Few-shot Novel View Synthesis. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 9065--9076. IEEE Computer Society

    2023

  22. [30]

    Wang, H.; Hu, W.; Zhu, L.; and Lau, R. W. 2024. Inverse Rendering of Glossy Objects via the Neural Plenoptic Function and Radiance Fields. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 19999--20008. IEEE Computer Society

    2024

  23. [31]

    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

  24. [32]

    Wu, S.; Basu, S.; Broedermann, T.; Van Gool, L.; and Sakaridis, C. 2025. PBR-NeRF : Inverse Rendering with Physics-Based Neural Fields. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE Computer Society

    2025

  25. [33]

    Yan, C.; Qu, D.; Xu, D.; Zhao, B.; Wang, Z.; Wang, D.; and Li, X. 2024. GS-SLAM: Dense Visual SLAM with 3D Gaussian Splatting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 19595--19604. IEEE Computer Society

    2024

  26. [34]

    Yang, W.; Chen, G.; Chen, C.; Chen, Z.; and Wong, K.-Y. K. 2022. PS-NeRF: Neural Inverse Rendering for Multi-view Photometric Stereo. In European conference on computer vision, 266--284. Springer

    2022

  27. [35]

    Yang, Z.; Gao, X.; Zhou, W.; Jiao, S.; Zhang, Y.; and Jin, X. 2024. Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene Reconstruction. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 20331--20341. IEEE Computer Society

    2024

  28. [36]

    Yao, Y.; Zeng, Z.; Gu, C.; Zhu, X.; and Zhang, L. 2025. Reflective Gaussian Splatting. In Processings of International Conference on Representation Learning, volume 2025, 68695--68711

    2025

  29. [37]

    Yao, Y.; Zhang, J.; Liu, J.; Qu, Y.; Fang, T.; McKinnon, D.; Tsin, Y.; and Quan, L. 2022. Neilf: Neural incident light field for physically-based material estimation. In European conference on computer vision, 700--716. Springer

    2022

  30. [38]

    Zhang, J.; Yao, Y.; Li, S.; Liu, J.; Fang, T.; McKinnon, D.; Tsin, Y.; and Quan, L. 2023. NeILF++: Inter-Reflectable Light Fields for Geometry and Material Estimation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 3601--3610. IEEE Computer Society

    2023

  31. [39]

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

    Zhang, R.; Isola, P.; Efros, A. A.; Shechtman, E.; and Wang, O. 2018. The Unreasonable Effectiveness of Deep Features as a Perceptual Metric. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. IEEE Computer Society

    2018

  32. [40]

    P.; Deng, B.; Debevec, P.; Freeman, W

    Zhang, X.; Srinivasan, P. P.; Deng, B.; Debevec, P.; Freeman, W. T.; and Barron, J. T. 2021. NeRFactor: neural factorization of shape and reflectance under an unknown illumination. ACM Trans. Graph., 40(6)

    2021

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.