pith. sign in

arxiv: 2606.08469 · v1 · pith:XWYNYSW6new · submitted 2026-06-07 · 💻 cs.GR · cs.CV

OctaOctree Neural Radiosity for Real-time Glossy Material Rendering

Jierui Ren , Haojie Jin , Bo Pang , Meng Gai , Fei Zhu , Yisong Chen , Sheng Li (Peking University) This is my paper

Pith reviewed 2026-06-27 17:52 UTC · model grok-4.3

classification 💻 cs.GR cs.CV
keywords OctaOctreeneural radiosityglobal illuminationglossy materialsradiance representationoctreeoctahedral mapreal-time rendering
0
0 comments X

The pith

OctaOctree combines an adaptive octree with octahedral directional maps to encode indirect illumination effects including sharp glossy reflections using a single network query.

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

This paper introduces OctaOctree as a spatial-angular representation for outgoing radiance in global illumination. An adaptive octree divides 3D space while each node holds an octahedral map for directions. The coupling allows fine spatial detail for local changes and richer angular resolution for glossy effects, embedding a prior that lets the neural model handle view-dependent radiance compactly. This leads to high-quality direction-aware illumination computed with one query at primary intersections and real-time speeds.

Core claim

OctaOctree organizes outgoing radiance with an adaptive octree in 3D space and associates each spatial node with an octahedral directional map. Coupling the spatial hierarchy with direction-dependent storage allocates fine spatial resolution to local illumination and visibility changes while using coarser spatial levels with richer angular resolution to capture glossy and specular radiance distributions. This embeds a reflectance-aware spatial-angular prior directly into the radiance representation, reducing the burden on neural networks to recover high-frequency view-dependent effects from positional features alone and providing a compact encoding for indirect illumination from diffuse to s

What carries the argument

The OctaOctree structure that pairs an adaptive 3D octree hierarchy with per-node octahedral directional maps to couple spatial and angular storage of radiance.

If this is right

  • High-quality direction-aware global illumination is produced with a single network query at primary intersections.
  • Real-time performance is achieved compared to baseline neural radiosity and radiance caching approaches.
  • Indirect illumination effects from diffuse interreflection to sharp glossy reflections are represented effectively.
  • The neural encoding becomes more compact by embedding the reflectance-aware prior in the representation structure.

Where Pith is reading between the lines

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

  • This representation could support extensions to dynamic scenes by updating the octree structure incrementally.
  • Similar spatial-angular hierarchies might apply to other transport problems such as volume rendering with view-dependent scattering.
  • The single-query property may enable its use in interactive applications like games where multiple bounces are needed without high cost.

Load-bearing premise

That the specific coupling of octree spatial nodes with octahedral maps embeds a reflectance-aware prior reducing the neural network's task of recovering high-frequency directional effects.

What would settle it

Measuring PSNR or SSIM error and milliseconds per frame on glossy material test scenes against a positional-encoding neural radiosity baseline to verify the claimed fidelity and speed improvements.

Figures

Figures reproduced from arXiv: 2606.08469 by Bo Pang, Fei Zhu, Haojie Jin, Jierui Ren, Meng Gai, Sheng Li (Peking University), Yisong Chen.

Figure 1
Figure 1. Figure 1: Two scenes rendered with our OctaOctree neural radiosity. Our method efficiently captures high-frequency view-dependent effects for glossy BSDF [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The overall pipeline of our method. We represent outgoing radiance [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: (a) An octahedron and corresponding unfolded map. (b) For a queried [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: The direction shift mechanism in our method. The predicted disparity [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Visual and qualitative comparisons on the main test scenes. We present the rendering results of our method ( [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: Visualization of ray-traced disparity and predicted disparity. The [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 6
Figure 6. Figure 6: Visualization of learned features at different OctaOctree levels. [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
Figure 9
Figure 9. Figure 9: Ablation of spatial and angular interpolation. Without interpolation, [PITH_FULL_IMAGE:figures/full_fig_p010_9.png] view at source ↗
read the original abstract

Modeling high-frequency outgoing radiance distributions remains a fundamental challenge in global illumination, especially for glossy and specular materials. Existing neural-based radiance caching methods commonly rely on positional feature encodings or spatially organized caches, which makes it difficult to represent sharp directional radiance variations without increasing the model complexity or sampling cost. To address this challenge, we propose OctaOctree, an efficient spatial-angular radiance representation for global illumination. OctaOctree organizes outgoing radiance with an adaptive octree in 3D space, and associates each spatial node with an octahedral directional map. By coupling the spatial hierarchy with direction-dependent storage, our representation allocates fine spatial resolution to local illumination and visibility changes, while using coarser spatial levels with richer angular resolution to capture glossy and specular radiance distributions. This design embeds a reflectance-aware spatial-angular prior directly into the radiance representation, reducing the burden on neural networks or reconstruction modules to recover high-frequency view-dependent effects from positional features alone. As a result, OctaOctree provides a compact and expressive neural encoding for a wide range of indirect illumination effects, from diffuse interreflection to sharp glossy reflections. Experiments demonstrate that our method produces high-quality, direction-aware global illumination with single network query at primary intersections, achieving improved fidelity and real-time performance compared with baseline neural radiosity and radiance caching approaches.

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

0 major / 2 minor

Summary. The paper proposes OctaOctree, a neural radiance representation for global illumination that couples an adaptive spatial octree hierarchy in 3D with per-node octahedral directional maps. This design is claimed to embed a reflectance-aware spatial-angular prior, enabling compact encoding of effects from diffuse interreflections to sharp glossy reflections. The method requires only a single network query at primary intersections and is reported to achieve improved fidelity and real-time performance over baseline neural radiosity and radiance caching approaches.

Significance. If the experimental claims hold, the approach could provide a practical advance in real-time glossy global illumination by reducing the modeling burden on neural networks through an explicit spatial-angular structure. The combination of octree adaptivity with octahedral storage is a concrete design choice that directly targets high-frequency directional variation without proportional increases in network capacity or sampling cost.

minor comments (2)
  1. The abstract states that experiments demonstrate improved fidelity and real-time performance, but the provided text contains no quantitative metrics, baseline comparisons, or scene descriptions. Adding a results section with specific numbers (e.g., PSNR, timing, memory) would strengthen the claims.
  2. The description of how the octahedral maps are queried and how the network is trained is high-level; explicit pseudocode or a diagram of the forward pass would clarify the single-query claim.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive evaluation of OctaOctree and for recommending minor revision. The report raises no specific major comments, so our response below is correspondingly brief. We will incorporate any minor editorial or presentation suggestions in the revised manuscript.

Circularity Check

0 steps flagged

No significant circularity identified

full rationale

The paper proposes OctaOctree as an architectural design coupling an adaptive octree with per-node octahedral directional maps to embed a reflectance-aware spatial-angular prior. No equations, fitted parameters, predictions, or derivation steps appear in the provided abstract or description. The central claim is a direct statement of the representation's properties and benefits, with no reduction of outputs to inputs by construction, no self-citation load-bearing premises, and no renaming of known results. Experiments are described as comparisons to baselines, leaving the method self-contained against external evaluation.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 1 invented entities

The abstract provides no information on free parameters, mathematical axioms, or other foundational elements; the contribution is described at a high level.

invented entities (1)
  • OctaOctree no independent evidence
    purpose: To organize outgoing radiance with adaptive octree in space and octahedral directional maps
    The representation is introduced in the paper as a new structure.

pith-pipeline@v0.9.1-grok · 5779 in / 1211 out tokens · 31526 ms · 2026-06-27T17:52:19.964855+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

55 extracted references · 38 canonical work pages

  1. [1]

    Steve Bako, Thijs Vogels, Brian McWilliams, Mark Meyer, Jan Novák, Alex Harvill, Pradeep Sen, Tony DeRose, and Fabrice Rousselle. 2017. Kernel-Predicting Con- volutional Networks for Denoising Monte Carlo Renderings.ACM Transactions on Graphics36, 4, Article 97 (2017). doi:10.1145/3072959.3073708

    work page doi:10.1145/3072959.3073708 2017

  2. [2]

    In: 2021 IEEE/CVF International Conference on Computer Vision (ICCV)

    Jonathan T. Barron, Ben Mildenhall, Matthew Tancik, Peter Hedman, Ricardo Martin-Brualla, and Pratul P. Srinivasan. 2021. Mip-NeRF: A Multiscale Represen- tation for Anti-Aliasing Neural Radiance Fields. InProceedings of the IEEE/CVF International Conference on Computer Vision. 5855–5864. doi:10.1109/ICCV48922. 2021.00580

    work page doi:10.1109/iccv48922 2021

  3. [3]

    Scalable diffusion models with transformers

    Jonathan T. Barron, Ben Mildenhall, Dor Verbin, Pratul P. Srinivasan, and Peter Hedman. 2023. Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 19697– 19705. doi:10.1109/ICCV51070.2023.01804

    work page doi:10.1109/iccv51070.2023.01804 2023

  4. [4]

    Alla Chaitanya, Anton S

    Chakravarty R. Alla Chaitanya, Anton S. Kaplanyan, Christoph Schied, Marco Salvi, Aaron Lefohn, Derek Nowrouzezahrai, and Timo Aila. 2017. Interactive Reconstruction of Monte Carlo Image Sequences using a Recurrent Denoising Autoencoder.ACM Transactions on Graphics36, 4, Article 98 (2017). doi:10.1145/ 3072959.3073601

    arXiv 2017

  5. [5]

    Anpei Chen, Zexiang Xu, Andreas Geiger, Jingyi Yu, and Hao Su. 2022. TensoRF: Tensorial Radiance Fields. InEuropean Conference on Computer Vision. 333–350. doi:10.1007/978-3-031-19824-3_20

    work page doi:10.1007/978-3-031-19824-3_20 2022

  6. [6]

    Dominici, Christian Döring, Joerg H

    Arno Coomans, Edoardo A. Dominici, Christian Döring, Joerg H. Mueller, Jozef Hladky, and Markus Steinberger. 2024. Real-time Neural Rendering of Dynamic Light Fields.Computer Graphics Forum43, 2 (2024). doi:10.1111/cgf.15014

    work page doi:10.1111/cgf.15014 2024

  7. [7]

    Mikhail Dereviannykh, Dmitrii Klepikov, Johannes Hanika, and Carsten Dachs- bacher. 2025. Neural Two-Level Monte Carlo Real-Time Rendering.Computer Graphics Forum44, 2 (2025). doi:10.1111/cgf.70050

    work page doi:10.1111/cgf.70050 2025

  8. [8]

    Stavros Diolatzis, Julien Philip, and George Drettakis. 2022. Active Exploration for Neural Global Illumination of Variable Scenes.ACM Transactions on Graphics 41, 4, Article 129 (2022)

    2022

  9. [9]

    Honghao Dong, Rui Su, Guoping Wang, and Sheng Li. 2024. Efficient Neural Path Guiding with 4D Modeling. InSIGGRAPH Asia 2024 Conference Papers. doi:10.1145/3680528.3687687

    work page doi:10.1145/3680528.3687687 2024

  10. [10]

    Honghao Dong, Guoping Wang, and Sheng Li. 2023. Neural Parametric Mixtures for Path Guiding. InSIGGRAPH ’23 Conference Proceedings. doi:10.1145/3588432. 3591533

    work page doi:10.1145/3588432 2023

  11. [11]

    Thomas Engelhardt and Carsten Dachsbacher. 2008. Octahedron Environment Maps. InVision, Modeling, and Visualization. 343–349

    2008

  12. [12]

    Sara Fridovich-Keil, Giacomo Meanti, Frederik Rahbæk Warburg, Benjamin Recht, and Angjoo Kanazawa. 2023. K-Planes: Explicit Radiance Fields in Space, Time, and Appearance. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 12479–12488. doi:10.1109/CVPR52729.2023.01201

    work page doi:10.1109/cvpr52729.2023.01201 2023

  13. [13]

    Sara Fridovich-Keil, Alex Yu, Matthew Tancik, Qinhong Chen, Benjamin Recht, and Angjoo Kanazawa. 2022. Plenoxels: Radiance Fields without Neural Net- works. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 5501–5510. doi:10.1109/CVPR52688.2022.00542

    work page doi:10.1109/cvpr52688.2022.00542 2022

  14. [14]

    Duan Gao, Haoyuan Mu, and Kun Xu. 2023. Neural Global Illumination: In- teractive Indirect Illumination Prediction Under Dynamic Area Lights.IEEE Transactions on Visualization and Computer Graphics29, 12 (2023), 5325–5341. doi:10.1109/TVCG.2022.3209963

    work page doi:10.1109/tvcg.2022.3209963 2023

  15. [15]

    Jonathan Granskog, Fabrice Rousselle, Marios Papas, and Jan Novák. 2020. Com- positional Neural Scene Representations for Shading Inference.ACM Transactions on Graphics39, 4, Article 137 (2020). doi:10.1145/3386569.3392405

    work page doi:10.1145/3386569.3392405 2020

  16. [16]

    Jie Guo, Zijing Zong, Yadong Song, Xihao Fu, Chengzhi Tao, Yanwen Guo, and Ling-Qi Yan. 2022. Efficient Light Probes for Real-Time Global Illumination.ACM Transactions on Graphics41, 6, Article 202 (2022). doi:10.1145/3550454.3555452

    work page doi:10.1145/3550454.3555452 2022

  17. [17]

    Yuan-Chen Guo, Di Kang, Linchao Bao, Yu He, and Song-Hai Zhang. 2022. NeR- FReN: Neural Radiance Fields with Reflections. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 18409–18418. doi:10.1109/ CVPR52688.2022.01786

    arXiv 2022

  18. [18]

    Saeed Hadadan, Karol Myszkowski, Hans-Peter Seidel, Tobias Ritschel, and Martin Weier. 2021. Neural Radiosity.ACM Transactions on Graphics40, 6, Article 215 (2021). doi:10.1145/3478513.3480491

    work page doi:10.1145/3478513.3480491 2021

  19. [19]

    Wenbo Hu, Yuling Wang, Lin Ma, Bangbang Yang, Lin Gao, Xiao Liu, and Yuewen Ma. 2023. Tri-MipRF: Tri-Mip Representation for Efficient Anti-Aliasing Neu- ral Radiance Fields. InProceedings of the IEEE/CVF International Conference on 10•Jierui Ren, Haojie Jin, Bo Pang, Meng Gai, Fei Zhu, Yisong Chen, and Sheng Li ReferenceFull modelRT depthNo dir.shiftOutgo...

    work page doi:10.1109/iccv51070.2023.01811 2023

  20. [20]

    Mustafa Işık, Krishna Mullia, Matthew Fisher, Jonathan Eisenmann, and Michaël Gharbi. 2021. Interactive Monte Carlo Denoising using Affinity of Neural Features. ACM Transactions on Graphics40, 4, Article 37 (2021). doi:10.1145/3450626.3459793

    work page doi:10.1145/3450626.3459793 2021

  21. [21]

    Lingjie Liu, Jiatao Gu, Kyaw Zaw Lin, Tat-Seng Chua, and Christian Theobalt

  22. [22]

    InAdvances in Neural Information Processing Systems, Vol

    Neural Sparse Voxel Fields. InAdvances in Neural Information Processing Systems, Vol. 33. 15651–15663

  23. [23]

    Zander Majercik, Cyril Crassin, Peter Shirley, and Morgan McGuire. 2019. Dy- namic Diffuse Global Illumination with Ray-Traced Irradiance Fields.Journal of Computer Graphics Techniques8, 2 (2019), 1–30

    2019

  24. [24]

    Srinivasan, Matthew Tancik, Jonathan T

    Ben Mildenhall, Pratul P. Srinivasan, Matthew Tancik, Jonathan T. Barron, Ravi Ramamoorthi, and Ren Ng. 2021. NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis. InCommunications of the ACM, Vol. 65. 99–106. doi:10. 1145/3503250 ReferenceFull modelNo octreeNo octa. MAPELPIPS 0.0560.0720.0610.0870.0630.077 0.1420.2340.1490.2380.1380.218...

    arXiv 2021

  25. [25]

    Shaohua Mo, Chuankun Zheng, Zihao Lin, Dianbing Xi, Qi Ye, Rui Wang, Hujun Bao, and Yuchi Huo. 2025. Dual-Band Feature Fusion for Neural Global Illumi- nation with Multi-Frequency Reflections. InSIGGRAPH 2025 Conference Papers. Article 49. doi:10.1145/3721238.3730733

    work page doi:10.1145/3721238.3730733 2025

  26. [26]

    Thomas Müller, Alex Evans, Christoph Schied, and Alexander Keller. 2022. In- stant Neural Graphics Primitives with a Multiresolution Hash Encoding.ACM Transactions on Graphics41, 4, Article 102 (2022). doi:10.1145/3528223.3530127

    work page doi:10.1145/3528223.3530127 2022

  27. [27]

    Thomas Müller, Brian McWilliams, Fabrice Rousselle, Markus Gross, and Jan Novák. 2019. Neural Importance Sampling.ACM Transactions on Graphics38, 5, Article 145 (2019). doi:10.1145/3341156

    work page doi:10.1145/3341156 2019

  28. [28]

    Thomas Müller, Fabrice Rousselle, Alexander Keller, and Jan Novák. 2020. Neural Control Variates.ACM Transactions on Graphics39, 6, Article 243 (2020). doi:10. 1145/3414685.3417819

    arXiv 2020

  29. [29]

    Thomas Müller, Fabrice Rousselle, Jan Novák, and Alexander Keller. 2021. Real- Time Neural Radiance Caching for Path Tracing.ACM Transactions on Graphics 40, 4, Article 36 (2021). doi:10.1145/3450626.3459812

    work page doi:10.1145/3450626.3459812 2021

  30. [30]

    Oliver Nalbach, Elena Arabadzhiyska, Deepak Mehta, Hans-Peter Seidel, and Tobias Ritschel. 2017. Deep Shading: Convolutional Neural Networks for Screen- Space Shading.Computer Graphics Forum36, 4 (2017), 65–78. doi:10.1111/cgf. 13225

    work page doi:10.1111/cgf 2017

  31. [31]

    Christian Reiser, Songyou Peng, Yiyi Liao, and Andreas Geiger. 2021. KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs. InProceedings of the IEEE/CVF International Conference on Computer Vision. 14335–14345. doi:10. 1109/ICCV48922.2021.01407

    arXiv 2021

  32. [32]

    Haocheng Ren, Yuchi Huo, Yifan Peng, Hongtao Sheng, Weidong Xue, Hongxiang Huang, Jingzhen Lan, Rui Wang, and Hujun Bao. 2024. LightFormer: Light- Oriented Global Neural Rendering in Dynamic Scene.ACM Transactions on OctaOctree Neural Radiosity for Real-time Glossy Material Rendering•11 Graphics(2024). doi:10.1145/3658229

    work page doi:10.1145/3658229 2024

  33. [33]

    Jierui Ren, Haojie Jin, Bo Pang, Yisong Chen, Guoping Wang, and Sheng Li. 2025. Neural Cone Radiosity for Interactive Global Illumination with Glossy Materials. arXiv preprint arXiv:2509.07522(2025). doi:10.48550/arXiv.2509.07522

    work page doi:10.48550/arxiv.2509.07522 2025

  34. [34]

    Juan Canés Rodriguez, Manuel Thomas, Tomáš Davidovič, Tobias Ritschel, Karol Myszkowski, Hans-Peter Seidel, Thorsten Grosch, Jaroslav Křivánek, and Martin Weier. 2020. Glossy Probe Reprojection for Interactive Global Illumination.ACM Transactions on Graphics39, 4, Article 122 (2020). doi:10.1145/3386569.3392483

    work page doi:10.1145/3386569.3392483 2020

  35. [35]

    Alla Chaitanya, John Burgess, Shiqiu Liu, Carsten Dachsbacher, Aaron Lefohn, and Marco Salvi

    Christoph Schied, Anton Kaplanyan, Chris Wyman, Anjul Patney, Chakravarty R. Alla Chaitanya, John Burgess, Shiqiu Liu, Carsten Dachsbacher, Aaron Lefohn, and Marco Salvi. 2017. Spatiotemporal Variance-Guided Filtering: Real-Time Reconstruction for Path-Traced Global Illumination. InProceedings of High Per- formance Graphics. 1–12. doi:10.1145/3105762.3105770

    work page doi:10.1145/3105762.3105770 2017

  36. [36]

    Peter-Pike Sloan, Jan Kautz, and John Snyder. 2002. Precomputed Radiance Trans- fer for Real-Time Rendering in Dynamic, Low-Frequency Lighting Environments. ACM Transactions on Graphics21, 3 (2002), 527–536. doi:10.1145/566570.566612

    work page doi:10.1145/566570.566612 2002

  37. [37]

    Rui Su, Honghao Dong, Haojie Jin, Yisong Chen, Guoping Wang, and Sheng Li. 2025. Vertex Features for Neural Global Illumination. InProceedings of the SIGGRAPH Asia 2025 Conference Papers. 1–11

    2025

  38. [38]

    Rui Su, Honghao Dong, Jierui Ren, Haojie Jin, et al . 2024. Dynamic Neural Radiosity with Multi-grid Decomposition.ACM SIGGRAPH Asia 2024 Conference Papers(2024). doi:10.1145/3680528.3687685

    work page doi:10.1145/3680528.3687685 2024

  39. [39]

    Cheng Sun, Min Sun, and Hwann-Tzong Chen. 2022. Direct Voxel Grid Optimiza- tion: Super-fast Convergence for Radiance Fields Reconstruction. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 5450–5459. doi:10.1109/CVPR52688.2022.00538

    work page doi:10.1109/cvpr52688.2022.00538 2022

  40. [40]

    Srinivasan, Ben Mildenhall, Sara Fridovich-Keil, Nithin Raghavan, Utkarsh Singhal, Ravi Ramamoorthi, Jonathan T

    Matthew Tancik, Pratul P. Srinivasan, Ben Mildenhall, Sara Fridovich-Keil, Nithin Raghavan, Utkarsh Singhal, Ravi Ramamoorthi, Jonathan T. Barron, and Ren Ng

  41. [41]

    Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains.Advances in Neural Information Processing Systems33 (2020), 7537–7547

    2020

  42. [42]

    Jiaxiang Tang, Xiaokang Chen, Jingbo Wang, and Gang Zeng. 2022. Compressible- composable NeRF via Rank-residual Decomposition.Advances in Neural Informa- tion Processing Systems35 (2022), 14798–14809

    2022

  43. [43]

    Manuel Thomas and Angus G. Forbes. 2017. Deep Illumination: Approximating Dynamic Global Illumination with Generative Adversarial Networks.Computer Graphics Forum36, 8 (2017), 111–120. doi:10.1111/cgf.13333

    work page doi:10.1111/cgf.13333 2017

  44. [44]

    Ego4d: Around the world in 3, 000 hours of egocentric video

    Dor Verbin, Peter Hedman, Ben Mildenhall, Todd Zickler, Jonathan T. Barron, and Pratul P. Srinivasan. 2022. Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance Fields. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 5481–5490. doi:10.1109/CVPR52688.2022.00541

    work page doi:10.1109/cvpr52688.2022.00541 2022

  45. [45]

    Thijs Vogels, Fabrice Rousselle, Brian McWilliams, Mark Meyer, Jan Novák, Alex Harvill, Pradeep Sen, Tony DeRose, and Nima Khademi Kalantari. 2018. Denoising with Kernel Prediction and Asymmetric Loss Functions.ACM Transactions on Graphics37, 4, Article 124 (2018). doi:10.1145/3197517.3201388

    work page doi:10.1145/3197517.3201388 2018

  46. [46]

    Peng-Shuai Wang, Yang Liu, Yu-Xiao Guo, Chun-Yu Sun, and Xin Tong. 2017. O-cnn: Octree-based convolutional neural networks for 3d shape analysis.ACM Transactions On Graphics (TOG)36, 4 (2017), 1–11

    2017

  47. [47]

    Songyin Wu, Sungye Kim, Zheng Zeng, Deepak Vembar, Sangeeta Jha, Anton Kaplanyan, and Ling-Qi Yan. 2023. ExtraSS: A Framework for Joint Spatial Super Sampling and Frame Extrapolation. InSIGGRAPH Asia 2023 Conference Papers. Article 92. doi:10.1145/3610548.3618224

    work page doi:10.1145/3610548.3618224 2023

  48. [48]

    Jia Xin, Michael McCool, Kayvon Fatahalian, Stephen Hill, Aaron Lefohn, and Jonathan Ragan-Kelley. 2022. Lightweight Bilateral Convolutional Neural Networks for Interactive Single-Bounce Diffuse Indirect Illumination.IEEE Transactions on Visualization and Computer Graphics28, 4 (2022), 1911–1923. doi:10.1109/TVCG.2020.3023129

    work page doi:10.1109/tvcg.2020.3023129 2022

  49. [49]

    Kun Xu, Yan-Pei Cao, Li-Qian Ma, Zhao Dong, Rui Wang, and Shi-Min Hu. 2014. A Practical Algorithm for Rendering Interreflections with All-frequency BRDFs. ACM Transactions on Graphics33, 1, Article 10 (2014). doi:10.1145/2533687

    work page doi:10.1145/2533687 2014

  50. [50]

    Alex Yu, Ruilong Li, Matthew Tancik, Hao Li, Ren Ng, and Angjoo Kanazawa. 2021. PlenOctrees for Real-Time Rendering of Neural Radiance Fields. InProceedings of the IEEE/CVF International Conference on Computer Vision. 5752–5761

    2021

  51. [51]

    Efros, Eli Shechtman, and Oliver Wang

    Richard Zhang, Phillip Isola, Alexei A. Efros, Eli Shechtman, and Oliver Wang

  52. [52]

    In: 2018 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2018, Salt Lake City, UT, USA, June 18-22, 2018

    The Unreasonable Effectiveness of Deep Features as a Perceptual Metric. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 586–595. doi:10.1109/CVPR.2018.00068

    work page doi:10.1109/cvpr.2018.00068 2018

  53. [53]

    Chuankun Zheng, Yuchi Huo, Hongxiang Huang, Hongtao Sheng, Junrong Huang, Rui Tang, Hao Zhu, Rui Wang, and Hujun Bao. 2024. Neural Global Illumination via Superposed Deformable Feature Fields. InSIGGRAPH Asia 2024 Conference Papers. doi:10.1145/3680528.3687680

    work page doi:10.1145/3680528.3687680 2024

  54. [54]

    Chuankun Zheng, Yuchi Huo, Shaohua Mo, Zhihua Zhong, et al . 2023. NeLT: Object-oriented Neural Light Transfer.ACM Transactions on Graphics42, 5, Article 152 (2023). doi:10.1145/3596491

    work page doi:10.1145/3596491 2023

  55. [55]

    Zhihua Zhong, Jingsen Zhu, Yuxin Dai, Chuankun Zheng, Guanlin Chen, Yuchi Huo, Hujun Bao, and Rui Wang. 2023. FuseSR: Super Resolution for Real-Time Rendering through Efficient Multi-resolution Fusion. InSIGGRAPH Asia 2023 Conference Papers. doi:10.1145/3610548.3618209

    work page doi:10.1145/3610548.3618209 2023