pith. sign in

arxiv: 2606.02292 · v1 · pith:Y6TVQ4JZnew · submitted 2026-06-01 · 💻 cs.CV

Neural Acquisition & Representation of Subsurface Scattering

Arjun Majumdar , Raphael Braun , Hendrik Lensch This is my paper

Pith reviewed 2026-06-28 15:22 UTC · model grok-4.3

classification 💻 cs.CV
keywords subsurface scatteringpixel footprintU-Netneural acquisitionrelightinglight transport3D scanninggeneralization
0
0 comments X

The pith

A U-Net CNN trained on 3D scans and projector-camera captures predicts pixel-level subsurface scattering responses that match real data and generalize to unseen materials.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper sets out to demonstrate that subsurface scattering properties can be acquired at high detail by learning the light footprint response at each surface point rather than modeling full volume transport. It feeds 3D scan data into a U-Net CNN whose training data come from efficient stereo projector-camera captures using phase-shifted patterns across multiple objects and views. If the approach holds, relighting becomes possible with any high-resolution projector pattern and the learned representations transfer to new scattering materials without retraining. A sympathetic reader would care because this replaces per-material physical simulation with a single learned mapping that produces nearly identical results to real illumination. The work therefore aims to make detailed subsurface effects practical for graphics and vision pipelines.

Core claim

Reconstructing dense pixel footprints via a U-Net CNN that takes 3D scans as input allows the model to produce relit color images whose footprints are almost identical to those measured from real-world projector illumination; training the same network on multiple views and objects yields representations that generalize directly to unseen subsurface scattering materials.

What carries the argument

U-Net CNN that maps 3D scanning input to dense pixel footprint responses for subsurface light transport

If this is right

  • Relighting with arbitrary high-resolution projector patterns becomes feasible from the predicted footprints.
  • The same trained model supports multiple views across multiple objects without per-object retraining.
  • Learned representations transfer to unseen subsurface scattering materials.
  • Qualitative and quantitative matches to real captured images confirm the footprints are faithful.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The footprint representation might serve as a compact intermediate for other inverse-rendering tasks that currently rely on full volumetric simulation.
  • Similar acquisition pipelines could be tested on dynamic or time-varying scattering if the capture setup is extended to video rates.
  • The approach implicitly suggests that many participating-media effects in graphics could be replaced by surface-footprint predictors trained once on diverse examples.

Load-bearing premise

The method assumes that 3D scanning combined with phase-shifted projector-camera patterns can supply training data sufficient for the network to learn accurate footprints across varied objects.

What would settle it

Apply the trained model to a new scattering material never seen in training, generate relit images under the same projector patterns, and compare them pixel-for-pixel to fresh real captures; large systematic differences would falsify the generalization claim.

Figures

Figures reproduced from arXiv: 2606.02292 by Arjun Majumdar, Hendrik Lensch, Raphael Braun.

Figure 1
Figure 1. Figure 1: 3-step phase-shifted proliferometry (PSP) ex￾ample. The same pattern can be used to measure dispar￾ity as well as for recovering spatially varying scattering footprints. at least once. The pixel distance 𝑚 needs to be chosen such that the observed footprints clearly do not overlap. If 𝑚 is small, i.e., 40 pixels or less, then the light from the neighboring adjacent coordinates might affect the response for… view at source ↗
Figure 2
Figure 2. Figure 2: Camera captured target ground truth. As training data we acquire densely sampled scattering footprints by illuminating individual dots with a shift￾ing grid. The camera captured images are in raw, mosaiced, © 2025 The Author(s). Proceedings published by Eurographics - The European Association for Computer Graphics [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Attention U-Net data flow. Given horizontal and vertical PSP pattern around camera pixel 𝑥, 𝑦 as input, the U-Net predicts the corresponding anisotropic footprint. We remove the background that did not receive suf￾ficient projector illumination. For all remaining camera pixels, the U-Net predicts a separate footprint by crop￾ping the corresponding input tiles around the pixel. For relighting, the unwrapped… view at source ↗
Figure 4
Figure 4. Figure 4: (a) green-white (b) red-white (c) gradient [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 7
Figure 7. Figure 7: Comparison: Target & Relit Images - Leaf front view Object+View Pattern MSE∗ PSNR SSIM LPIPS∗ Soap-front red-white 6.14 52.00 0.995 5.31 Soap-front green-white 8.42 49.39 0.994 7.84 Soap-front gradient 5.75 47.82 0.999 2.84 Soap-back red-white 9.52 50.02 0.996 4.85 Soap-back green-white 6.22 52.04 0.995 7.41 Soap-back gradient 4.70 46.44 0.999 2.18 Orange-front red-white 1.62 46.52 0.997 2.18 Orange-front … view at source ↗
Figure 5
Figure 5. Figure 5: Comparison: Target & Relit Images - Soap front view (first two columns), back view (last two columns). Note how the smooth transition between dif￾ferent colors as well as the indents on the front are accurately reproduced. (a) Target (b) Output (c) Target (d) Output [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Comparison: Target & Relit Images - Or￾ange front view (first two columns), back view (last two columns). Here, the color shift due to the orange peel as well as the peels structure are captured accurately. in [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 8
Figure 8. Figure 8: Comparison: Target & Relit Images - Unseen Soap2 front & back view. No images of this object had been used during training. 6. Conclusions To obtain a relightable representation for subsurface scattering objects, we present a novel technique that can predict the per-pixel spatially varying pixel foot￾prints from just six input images captured with high￾frequency horizontal and vertical sinusoidal patterns.… view at source ↗
read the original abstract

We present a method to acquire and estimate the sub-surface scattering properties of light transport at a highly detailed level by learning the pixel footprint response at each point on the object surface. The reconstruction leverages 3D scanning techniques as input to a U-Net CNN. A stereo projector-camera setup using phase-shifted profilometry (PSP) patterns efficiently captures the data for a variety of scattering objects. Reconstructing dense pixel footprints allows for relighting with arbitrary high-resolution projector patterns. The final output is a relit color image. Qualitative and quantitative comparison against illuminated real-world captured images demonstrate that the predicted footprints are almost identical to the actual responses. The same model is trained for multiple views across multiple objects such that the learned representations can be used to generalize to unseen sub-surface scattering materials as well.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The paper presents a neural method to acquire and represent subsurface scattering by learning per-pixel footprint responses of light transport. It takes 3D scans as input to a U-Net CNN, captures training data via a stereo projector-camera system using phase-shifted profilometry (PSP) patterns, and produces relit images under arbitrary high-resolution projector patterns. The model is trained across multiple views and objects to enable generalization to unseen scattering materials, with the central empirical claim that predicted footprints are almost identical to real captured responses.

Significance. If the quantitative comparisons and generalization results hold with rigorous metrics, the work would offer a practical data-driven pipeline for high-detail subsurface scattering acquisition and relighting, potentially reducing reliance on hand-crafted physical models in computer graphics and vision applications.

major comments (2)
  1. [Abstract] Abstract: the claim that 'predicted footprints are almost identical to the actual responses' and that quantitative comparisons were performed is unsupported by any reported metrics, error bars, or description of the comparison protocol; this is load-bearing for the central claim and must be addressed with concrete numbers and methodology in the results section.
  2. The reconstruction pipeline (3D scan input to U-Net, PSP capture) is described at a high level but lacks details on loss functions, training data statistics, or how the pixel footprints are represented and decoded for relighting; these elements are necessary to assess reproducibility and whether the generalization claim is supported.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. The two major comments highlight areas where the manuscript can be strengthened for clarity and rigor. We address each point below and will revise accordingly.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the claim that 'predicted footprints are almost identical to the actual responses' and that quantitative comparisons were performed is unsupported by any reported metrics, error bars, or description of the comparison protocol; this is load-bearing for the central claim and must be addressed with concrete numbers and methodology in the results section.

    Authors: We agree that the abstract claim requires explicit quantitative support. The manuscript states that quantitative comparisons were performed against captured images, but the current version does not report specific metrics, error bars, or the evaluation protocol in sufficient detail. In the revised manuscript we will add these elements to the results section, including concrete numbers (e.g., mean absolute error or Pearson correlation between predicted and measured footprints), standard deviations across multiple test samples, and a clear description of the comparison protocol. revision: yes

  2. Referee: The reconstruction pipeline (3D scan input to U-Net, PSP capture) is described at a high level but lacks details on loss functions, training data statistics, or how the pixel footprints are represented and decoded for relighting; these elements are necessary to assess reproducibility and whether the generalization claim is supported.

    Authors: We acknowledge that the methods section provides only a high-level description. The revised manuscript will expand this section to include: (1) the exact loss function used for training the U-Net, (2) training data statistics (number of objects, views, and total footprint samples), and (3) the representation of pixel footprints and the decoding procedure that enables relighting under arbitrary high-resolution projector patterns. These additions will directly support assessment of reproducibility and the generalization results. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper presents an empirical acquisition and learning pipeline: PSP patterns and 3D scans are captured as input, a U-Net is trained to predict pixel footprints, and outputs are validated by direct comparison to separately captured real illuminated images. No derivation, equation, or loss is shown that reduces a claimed prediction to a fitted parameter or self-citation by construction. Generalization claims rest on held-out objects and views rather than re-using the training targets as the evaluation metric. The argument is therefore self-contained against external captured data.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the effectiveness of the data capture setup and the generalization ability of the trained model, with no explicit free parameters mentioned in the abstract.

axioms (1)
  • domain assumption U-Net architecture can learn the mapping from 3D scans to pixel footprints
    The paper relies on the standard assumption that CNNs can approximate the light transport function.

pith-pipeline@v0.9.1-grok · 5656 in / 1114 out tokens · 21901 ms · 2026-06-28T15:22:06.221228+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

81 extracted references · 14 canonical work pages

  1. [1]

    Delio Vicini and Vladlen Koltun and Wenzel Jakob , title =

  2. [3]

    Journal of the Optical Society of America A , volume =

    Simeon Geiger and Philipp Hank and Alwin Kienle , title =. Journal of the Optical Society of America A , volume =

  3. [4]

    Per Christensen and Julian Fong and Jonathan Shade and Wayne Wooten and Brenden Schubert and Andrew Kensler and Stephen Friedman and Charlie Kilpatrick and Cliff Ramshaw and Marc Bannister and Brenton Rayner and Jonathan Brouillat and Max Liani , title =

  4. [5]

    Luca Fascione and Johannes Hanika and Mark Leone and Marc Droske and Jorge Schwarzhaupt and Tomáš Davidovič and Andrea Weidlich and Johannes Meng , title =

  5. [6]

    and Welling, Max , biburl =

    Kingma, Diederik P. and Welling, Max , biburl =. Auto-Encoding Variational Bayes , added-at =. 2nd International Conference on Learning Representations,. http://arxiv.org/abs/1312.6114v10 , eprintclass =

    Pith/arXiv arXiv

  6. [7]

    Learning Structured Output Representation using Deep Conditional Generative Models , url =

    Sohn, Kihyuk and Lee, Honglak and Yan, Xinchen , booktitle =. Learning Structured Output Representation using Deep Conditional Generative Models , url =

  7. [8]

    Che and F

    C. Che and F. Luan and S. Zhao and K. Bala and I. Gkioulekas , title =. IEEE International Conference on Computational Photography (ICCP) , DOI =

  8. [9]

    U-Net: Convolutional Networks for Biomedical Image Segmentation

    Ronneberger, Olaf and Fischer, Philipp and Brox, Thomas , biburl =. U-Net: Convolutional Networks for Biomedical Image Segmentation. , url =. CoRR , keywords =

  9. [12]

    Optics and Lasers in Engineering , volume =

    Chao Zuo and Lei Huang and Minliang Zhang and Qian Chen and Anand Asundi , title =. Optics and Lasers in Engineering , volume =. doi:Optics and Lasers in Engineering , year =

  10. [13]

    Optics and Lasers in Engineering , DOI =

    Xiaoyu He and Qian Kemao , title =. Optics and Lasers in Engineering , DOI =

  11. [14]

    Chen, Tongbo and Seidel, Hans-Peter and Lensch, Hendrik P. A. , booktitle=. Modulated phase-shifting for 3D scanning , year=

  12. [15]

    Chen, Tongbo and Lensch, Hendrik P. A. and Fuchs, Christian and Seidel, Hans-Peter , booktitle=. Polarization and Phase-Shifting for 3D Scanning of Translucent Objects , year=

  13. [16]

    , booktitle=

    Gupta, Mohit and Nayar, Shree K. , booktitle=. Micro Phase Shifting , year=

  14. [17]

    Embedded phase shifting: Robust phase shifting with embedded signals , year=

    Moreno, Daniel and Son, Kilho and Taubin, Gabriel , booktitle=. Embedded phase shifting: Robust phase shifting with embedded signals , year=

  15. [18]

    and van Dam, Andries and Feiner, Stephen K

    Foley, James D. and van Dam, Andries and Feiner, Stephen K. and Hughes, John F. and Phillips, R. , TITLE =

  16. [20]

    Proceedings of the 34th International Conference on Neural Information Processing Systems , articleno =

    Ho, Jonathan and Jain, Ajay and Abbeel, Pieter , title =. Proceedings of the 34th International Conference on Neural Information Processing Systems , articleno =. 2020 , isbn =

    2020

  17. [21]

    Medical Imaging with Deep Learning , year=

    Attention U-Net: Learning Where to Look for the Pancreas , author=. Medical Imaging with Deep Learning , year=

  18. [22]

    and Li-wei He and Cutler, R

    Malvar, H.S. and Li-wei He and Cutler, R. , booktitle=. High-quality linear interpolation for demosaicing of Bayer-patterned color images , year=

  19. [23]

    Demosaicing With Directional Filtering and a posteriori Decision , year=

    Menon, Daniele and Andriani, Stefano and Calvagno, Giancarlo , journal=. Demosaicing With Directional Filtering and a posteriori Decision , year=

  20. [24]

    2011 , note =

    Getreuer, Pascal , journal =. 2011 , note =

    2011

  21. [25]

    Embrace the Gap: VAEs Perform Independent Mechanism Analysis , url =

    Reizinger, Patrik and Gresele, Luigi and Brady, Jack and von K\". Embrace the Gap: VAEs Perform Independent Mechanism Analysis , url =. Advances in Neural Information Processing Systems , editor =

  22. [26]

    Don t Blame the ELBO! A Linear VAE Perspective on Posterior Collapse , url =

    Lucas, James and Tucker, George and Grosse, Roger B and Norouzi, Mohammad , booktitle =. Don t Blame the ELBO! A Linear VAE Perspective on Posterior Collapse , url =

  23. [27]

    Posterior Collapse and Latent Variable Non-identifiability , url =

    Wang, Yixin and Blei, David and Cunningham, John P , booktitle =. Posterior Collapse and Latent Variable Non-identifiability , url =

  24. [28]

    The Usual Suspects?

    Dai, Bin and Wang, Ziyu and Wipf, David , booktitle =. The Usual Suspects?. 2020 , editor =

    2020

  25. [29]

    Biomedical Imaging and Sensing Conference , volume=

    Development of measurement system for subsurface scattering light of skin and analysis of its age-related changes , author=. Biomedical Imaging and Sensing Conference , volume=. 2023 , organization=

    2023

  26. [30]

    Proceedings of the 27th annual conference on Computer graphics and interactive techniques , pages=

    Acquiring the reflectance field of a human face , author=. Proceedings of the 27th annual conference on Computer graphics and interactive techniques , pages=

  27. [31]

    Proceedings of the 14th Eurographics workshop on Rendering , pages=

    Wavelet environment matting , author=. Proceedings of the 14th Eurographics workshop on Rendering , pages=

  28. [32]

    ACM Transactions on Graphics (ToG) , volume=

    Compressive light transport sensing , author=. ACM Transactions on Graphics (ToG) , volume=. 2009 , publisher=

    2009

  29. [34]

    , author=

    Primal-dual coding to probe light transport. , author=. ACM Trans. Graph. , volume=

  30. [35]

    Garg, Gaurav and Talvala, Eino-Ville and Levoy, Marc and Lensch, Hendrik P. A. , year =. Symposium on Rendering , editor =

  31. [37]

    ACM Transactions on Graphics (TOG) , volume=

    Neural light transport for relighting and view synthesis , author=. ACM Transactions on Graphics (TOG) , volume=. 2021 , publisher=

    2021

  32. [38]

    ACM Transactions on Graphics (ToG) , volume=

    Image based relighting using neural networks , author=. ACM Transactions on Graphics (ToG) , volume=. 2015 , publisher=

    2015

  33. [40]

    , author=

    Optical computing for fast light transport analysis. , author=. ACM Trans. Graph. , volume=. 2010 , publisher=

    2010

  34. [41]

    Proceedings of the 28th annual conference on Computer graphics and interactive techniques , pages=

    A practical model for subsurface light transport , author=. Proceedings of the 28th annual conference on Computer graphics and interactive techniques , pages=

  35. [42]

    ACM Transactions on Graphics (ToG) , volume=

    Analysis of human faces using a measurement-based skin reflectance model , author=. ACM Transactions on Graphics (ToG) , volume=. 2006 , publisher=

    2006

  36. [43]

    The Thirty-eighth Annual Conference on Neural Information Processing Systems , year=

    Subsurface Scattering for Gaussian Splatting , author=. The Thirty-eighth Annual Conference on Neural Information Processing Systems , year=

  37. [44]

    Advances in Neural Information Processing Systems , volume=

    Neural relightable participating media rendering , author=. Advances in Neural Information Processing Systems , volume=

  38. [45]

    European Conference on Computer Vision , pages=

    Neural radiance transfer fields for relightable novel-view synthesis with global illumination , author=. European Conference on Computer Vision , pages=. 2022 , organization=

    2022

  39. [47]

    Computer Graphics Forum , volume=

    Combining confocal imaging and descattering , author=. Computer Graphics Forum , volume=. 2008 , organization=

    2008

  40. [48]

    Applied optics , volume=

    Spatially resolved absolute diffuse reflectance measurements for noninvasive determination of the optical scattering and absorption coefficients of biological tissue , author=. Applied optics , volume=. 1996 , publisher=

    1996

  41. [49]

    : Renderman: An advanced path-tracing architecture for movie rendering

    Christensen P., Fong J., Shade J., Wooten W., Schubert B., Kensler A., Friedman S., Kilpatrick C., Ramshaw C., Bannister M., Rayner B., Brouillat J., Liani M. : Renderman: An advanced path-tracing architecture for movie rendering. ACM TOG 37 (2018)

    2018

  42. [50]

    Chen T., Lensch H. P. A., Fuchs C., Seidel H.-P. : Polarization and phase-shifting for 3d scanning of translucent objects. In 2007 IEEE Conference on Computer Vision and Pattern Recognition (2007), pp. 1--8. https://doi.org/10.1109/CVPR.2007.383209 doi:10.1109/CVPR.2007.383209

    work page doi:10.1109/cvpr.2007.383209 2007

  43. [51]

    Chen T., Seidel H.-P., Lensch H. P. A. : Modulated phase-shifting for 3d scanning. In 2008 IEEE Conference on Computer Vision and Pattern Recognition (2008), pp. 1--8. https://doi.org/10.1109/CVPR.2008.4587836 doi:10.1109/CVPR.2008.4587836

    work page doi:10.1109/cvpr.2008.4587836 2008

  44. [52]

    : Acquiring the reflectance field of a human face

    Debevec P., Hawkins T., Tchou C., Duiker H.-P., Sarokin W., Sagar M. : Acquiring the reflectance field of a human face. In Proceedings of the 27th annual conference on Computer graphics and interactive techniques (2000), pp. 145--156

    2000

  45. [53]

    : Subsurface scattering for gaussian splatting

    Dihlmann J.-N., Majumdar A., Engelhardt A., Braun R., Lensch H. : Subsurface scattering for gaussian splatting. In The Thirty-eighth Annual Conference on Neural Information Processing Systems (2025)

    2025

  46. [54]

    : The usual suspects? R eassessing blame for VAE posterior collapse

    Dai B., Wang Z., Wipf D. : The usual suspects? R eassessing blame for VAE posterior collapse. In Proceedings of the 37th International Conference on Machine Learning (13--18 Jul 2020), III H. D., Singh A., (Eds.), vol. 119 of Proceedings of Machine Learning Research, PMLR, pp. 2313--2322. URL: https://proceedings.mlr.press/v119/dai20c.html

    2020

  47. [55]

    Fuchs C., Heinz M., Levoy M., Seidel H.-P., Lensch H. P. : Combining confocal imaging and descattering. In Computer Graphics Forum (2008), vol. 27, Wiley Online Library, pp. 1245--1253

    2008

  48. [56]

    : Manuka: A batch-shading architecture for spectral path tracing in movie production

    Fascione L., Hanika J., Leone M., Droske M., Schwarzhaupt J., Davidovič T., Weidlich A., Meng J. : Manuka: A batch-shading architecture for spectral path tracing in movie production. ACM TOG 37 (2018)

    2018

  49. [57]

    : Malvar-He-Cutler Linear Image Demosaicking

    Getreuer P. : Malvar-He-Cutler Linear Image Demosaicking . Image Processing On Line 1 (2011), 83--89. https://doi.org/10.5201/ipol.2011.g_mhcd

    work page doi:10.5201/ipol.2011.g_mhcd 2011

  50. [58]

    : Improved topography reconstruction of volume scattering objects using structured light

    Geiger S., Hank P., Kienle A. : Improved topography reconstruction of volume scattering objects using structured light. Journal of the Optical Society of America A 39 (2022)

    2022

  51. [59]

    Goesele M., Lensch H. P. A., Lang J., Fuchs C., Seidel H.-P. : Disco: acquisition of translucent objects. ACM Trans. Graph. 23, 3 (Aug. 2004), 835–844. URL: https://doi.org/10.1145/1015706.1015807, https://doi.org/10.1145/1015706.1015807 doi:10.1145/1015706.1015807

    work page doi:10.1145/1015706.1015807 2004

  52. [60]

    Gupta M., Nayar S. K. : Micro phase shifting. In 2012 IEEE Conference on Computer Vision and Pattern Recognition (2012), pp. 813--820. https://doi.org/10.1109/CVPR.2012.6247753 doi:10.1109/CVPR.2012.6247753

    work page doi:10.1109/cvpr.2012.6247753 2012

  53. [61]

    Garg G., Talvala E.-V., Levoy M., Lensch H. P. A. : Symmetric Photography: Exploiting Data-sparseness in Reflectance Fields . In Symposium on Rendering (2006), Akenine-Moeller T., Heidrich W., (Eds.), The Eurographics Association. https://doi.org/10.2312/EGWR/EGSR06/251-262 doi:10.2312/EGWR/EGSR06/251-262

    work page doi:10.2312/egwr/egsr06/251-262 2006

  54. [62]

    : A comparative study on temporal phase unwrapping methods in high-speed fringe projection profilometry

    He X., Kemao Q. : A comparative study on temporal phase unwrapping methods in high-speed fringe projection profilometry. Optics and Lasers in Engineering 142 (2021). https://doi.org/https://doi.org/10.1016/j.optlaseng.2021.106613 doi:https://doi.org/10.1016/j.optlaseng.2021.106613

    work page doi:10.1016/j.optlaseng.2021.106613 2021

  55. [63]

    W., Marschner S

    Jensen H. W., Marschner S. R., Levoy M., Hanrahan P. : A practical model for subsurface light transport. In Proceedings of the 28th Annual Conference on Computer Graphics and Interactive Techniques (New York, NY, USA, 2001), SIGGRAPH '01, Association for Computing Machinery, p. 511–518. URL: https://doi.org/10.1145/383259.383319, https://doi.org/10.1145/3...

    work page doi:10.1145/383259.383319 2001

  56. [64]

    W., Marschner S

    Jensen H. W., Marschner S. R., Levoy M., Hanrahan P. : A practical model for subsurface light transport. In Proceedings of the 28th annual conference on Computer graphics and interactive techniques (2001), pp. 511--518

    2001

  57. [65]

    : Development of measurement system for subsurface scattering light of skin and analysis of its age-related changes

    Kikuchi K., Katsuyama M., Shibata T., Hardeberg J., Yuasa T., Aizu Y. : Development of measurement system for subsurface scattering light of skin and analysis of its age-related changes. In Biomedical Imaging and Sensing Conference (2023), vol. 12608, SPIE, pp. 123--126

    2023

  58. [66]

    S., Hibst R., Steiner R., Wilson B

    Kienle A., Lilge L., Patterson M. S., Hibst R., Steiner R., Wilson B. C. : Spatially resolved absolute diffuse reflectance measurements for noninvasive determination of the optical scattering and absorption coefficients of biological tissue. Applied optics 35, 13 (1996), 2304--2314

    1996

  59. [67]

    C., Lo W.-Y., Doll \'a r P., Girshick R

    Kirillov A., Mintun E., Ravi N., Mao H., Rolland C., Gustafson L., Xiao T., Whitehead S., Berg A. C., Lo W.-Y., Doll \'a r P., Girshick R. : Segment anything. arXiv:2304.02643 (2023)

    Pith/arXiv arXiv 2023

  60. [68]

    B., Norouzi M

    Lucas J., Tucker G., Grosse R. B., Norouzi M. : Don t blame the elbo! a linear vae perspective on posterior collapse. In Advances in Neural Information Processing Systems (2019), Wallach H., Larochelle H., Beygelzimer A., d Alch\' e -Buc F., Fox E., Garnett R., (Eds.), vol. 32, Curran Associates, Inc. URL: https://proceedings.neurips.cc/paper_files/paper/...

    2019

  61. [69]

    : Neural radiance transfer fields for relightable novel-view synthesis with global illumination

    Lyu L., Tewari A., Leimk \"u hler T., Habermann M., Theobalt C. : Neural radiance transfer fields for relightable novel-view synthesis with global illumination. In European Conference on Computer Vision (2022), Springer, pp. 153--169

    2022

  62. [70]

    : Demosaicing with directional filtering and a posteriori decision

    Menon D., Andriani S., Calvagno G. : Demosaicing with directional filtering and a posteriori decision. IEEE Transactions on Image Processing 16, 1 (2007), 132--141. https://doi.org/10.1109/TIP.2006.884928 doi:10.1109/TIP.2006.884928

    work page doi:10.1109/tip.2006.884928 2007

  63. [71]

    : Embedded phase shifting: Robust phase shifting with embedded signals

    Moreno D., Son K., Taubin G. : Embedded phase shifting: Robust phase shifting with embedded signals. In 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2015), pp. 2301--2309. https://doi.org/10.1109/CVPR.2015.7298843 doi:10.1109/CVPR.2015.7298843

    work page doi:10.1109/cvpr.2015.7298843 2015

  64. [72]

    : High-quality linear interpolation for demosaicing of bayer-patterned color images

    Malvar H., wei He L., Cutler R. : High-quality linear interpolation for demosaicing of bayer-patterned color images. In 2004 IEEE International Conference on Acoustics, Speech, and Signal Processing (2004), vol. 3, pp. iii--485. https://doi.org/10.1109/ICASSP.2004.1326587 doi:10.1109/ICASSP.2004.1326587

    work page doi:10.1109/icassp.2004.1326587 2004

  65. [73]

    K., Krishnan G., Grossberg M

    Nayar S. K., Krishnan G., Grossberg M. D., Raskar R. : Fast separation of direct and global components of a scene using high frequency illumination. ACM Trans. Graph. 25, 3 (July 2006), 935–944. URL: https://doi.org/10.1145/1141911.1141977, https://doi.org/10.1145/1141911.1141977 doi:10.1145/1141911.1141977

    work page doi:10.1145/1141911.1141977 2006

  66. [74]

    O'Toole M., Kutulakos K. N. : Optical computing for fast light transport analysis. ACM Trans. Graph. 29, 6 (2010), 164

    2010

  67. [75]

    O'Toole M., Raskar R., Kutulakos K. N. : Primal-dual coding to probe light transport. ACM Trans. Graph. 31, 4 (2012), 39--1

    2012

  68. [76]

    L., Lee M., Heinrich M., Misawa K., Mori K., McDonagh S., Hammerla N

    Oktay O., Schlemper J., Folgoc L. L., Lee M., Heinrich M., Misawa K., Mori K., McDonagh S., Hammerla N. Y., Kainz B., Glocker B., Rueckert D. : Attention u-net: Learning where to look for the pancreas. In Medical Imaging with Deep Learning (2018). URL: https://openreview.net/forum?id=Skft7cijM

    2018

  69. [77]

    : Wavelet environment matting

    Peers P., Dutr \'e P. : Wavelet environment matting. In Proceedings of the 14th Eurographics workshop on Rendering (2003), pp. 157--166

    2003

  70. [78]

    K., Lamond B., Ghosh A., Matusik W., Ramamoorthi R., Debevec P

    Peers P., Mahajan D. K., Lamond B., Ghosh A., Matusik W., Ramamoorthi R., Debevec P. : Compressive light transport sensing. ACM Transactions on Graphics (ToG) 28, 1 (2009), 1--18

    2009

  71. [79]

    : U-net: Convolutional networks for biomedical image segmentation

    Ronneberger O., Fischer P., Brox T. : U-net: Convolutional networks for biomedical image segmentation. CoRR (2015). URL: http://dblp.uni-trier.de/db/journals/corr/corr1505.html

    2015

  72. [80]

    u gelgen J., Zietlow D., Sch\

    Reizinger P., Gresele L., Brady J., von K\" u gelgen J., Zietlow D., Sch\" o lkopf B., Martius G., Brendel W., Besserve M. : Embrace the gap: Vaes perform independent mechanism analysis. In Advances in Neural Information Processing Systems (2022), Koyejo S., Mohamed S., Agarwal A., Belgrave D., Cho K., Oh A., (Eds.), vol. 35, Curran Associates, Inc., pp. ...

    2022

  73. [81]

    R., Horowitz M., Levoy M., Lensch H

    Sen P., Chen B., Garg G., Marschner S. R., Horowitz M., Levoy M., Lensch H. P. A. : Dual photography. In ACM SIGGRAPH 2005 Papers (New York, NY, USA, 2005), SIGGRAPH '05, Association for Computing Machinery, p. 745–755. URL: https://doi.org/10.1145/1186822.1073257, https://doi.org/10.1145/1186822.1073257 doi:10.1145/1186822.1073257

    work page doi:10.1145/1186822.1073257 2005

  74. [82]

    : A learned shape-adaptive subsurface scattering model

    Vicini D., Koltun V., Jakob W. : A learned shape-adaptive subsurface scattering model. ACM TOG 38 (2019)

    2019

  75. [83]

    Wang Y., Blei D., Cunningham J. P. : Posterior collapse and latent variable non-identifiability. In Advances in Neural Information Processing Systems (2021), Ranzato M., Beygelzimer A., Dauphin Y., Liang P., Vaughan J. W., (Eds.), vol. 34, Curran Associates, Inc., pp. 5443--5455. URL: https://proceedings.neurips.cc/paper_files/paper/2021/file/2b6921f2c64d...

    2021

  76. [84]

    W., et al

    Weyrich T., Matusik W., Pfister H., Bickel B., Donner C., Tu C., McAndless J., Lee J., Ngan A., Jensen H. W., et al. : Analysis of human faces using a measurement-based skin reflectance model. ACM Transactions on Graphics (ToG) 25, 3 (2006), 1013--1024

    2006

  77. [85]

    R., Gao R., Funkhouser T., Wu J

    Yu H.-X., Guo M., Fathi A., Chang Y.-Y., Chan E. R., Gao R., Funkhouser T., Wu J. : Learning object-centric neural scattering functions for free-viewpoint relighting and scene composition. arXiv preprint arXiv:2303.06138 (2023)

    arXiv 2023

  78. [86]

    : Phase shifting algorithms for fringe projection profilometry: A review

    Zuo C., Feng S., Huang L., Tao T., Yin W., Chen Q. : Phase shifting algorithms for fringe projection profilometry: A review. Optics and Lasers in Engineering 109 (2018). https://doi.org/doi:10.1016/j.optlaseng.2018.04.019 doi:doi:10.1016/j.optlaseng.2018.04.019

    work page doi:10.1016/j.optlaseng.2018.04.019 2018

  79. [87]

    : Temporal phase unwrapping algorithms for fringe projection profilometry: A comparative review

    Zuo C., Huang L., Zhang M., Chen Q., Asundi A. : Temporal phase unwrapping algorithms for fringe projection profilometry: A comparative review. Optics and Lasers in Engineering 85 (2016). https://doi.org/Optics and Lasers in Engineering doi:Optics and Lasers in Engineering

    2016

  80. [88]

    : Neural relighting with subsurface scattering by learning the radiance transfer gradient

    Zhu S., Saito S., Bozic A., Aliaga C., Darrell T., Lassner C. : Neural relighting with subsurface scattering by learning the radiance transfer gradient. arXiv preprint arXiv:2306.09322 (2023)

    arXiv 2023

Showing first 80 references.