pith. sign in

arxiv: 2605.19304 · v1 · pith:7M4KMGBOnew · submitted 2026-05-19 · 💻 cs.CV · cs.GR

MMGS: 10times Compressed 3DGS through Optimal Transport Aggregation based on Multi-view Ranking

Beizhen Zhao , Sicheng Yu , Ziran Yin , Dongxu Shen , Hao Wang This is my paper

Pith reviewed 2026-05-20 07:09 UTC · model grok-4.3

classification 💻 cs.CV cs.GR
keywords 3D Gaussian Splattingmodel compressionoptimal transportmulti-view consistency3D reconstructionrendering accelerationprimitive aggregation
0
0 comments X

The pith

Multi-view geometric ranking plus optimal transport aggregation compresses 3D Gaussian Splatting to 10 percent of its primitives while matching original rendering quality and training ten times faster.

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

The paper presents a compression framework for 3D Gaussian Splatting that treats the entire set of primitives as a global geometric distribution to be matched rather than pruned locally. It first ranks each Gaussian by its contribution across multiple views using geometric consistency, then applies optimal transport to merge redundant ones into fewer representatives that still reproduce the scene's light field. An additional OT-based densification step keeps the remaining Gaussians well-behaved during optimization. If the approach holds, 3D reconstruction pipelines could run with dramatically lower memory and compute budgets yet deliver comparable visual fidelity.

Core claim

By formulating Gaussian optimization as a global geometric distribution matching problem, the method integrates a multi-view contribution ranking step that filters primitives via geometric consistency, a global OT-based aggregation algorithm that merges redundancies while preserving underlying geometry, and an OT-based densification operator that maintains distributional properties; this combination yields state-of-the-art rendering quality using only 10 percent of the primitives and achieves 10 times accelerated training speeds relative to vanilla 3DGS.

What carries the argument

The multi-view 3D Gaussian contribution ranking mechanism, which scores each primitive by its geometric consistency across views, paired with a global Optimal Transport aggregation algorithm that solves a distribution matching problem to merge redundant primitives without distorting scene geometry.

If this is right

  • Scene models occupy roughly one-tenth the storage while delivering competitive PSNR and SSIM scores.
  • Optimization converges in approximately one-tenth the iterations because far fewer primitives are updated each step.
  • The OT aggregation replaces heuristic pruning thresholds with a global transport plan that respects the full scene distribution.
  • The same OT operator can be reused for controlled densification that keeps the Gaussian set well-conditioned.

Where Pith is reading between the lines

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

  • The same ranking-plus-transport pattern could be tested on other explicit scene representations such as point clouds or surfels to see whether similar compression ratios appear.
  • If the geometric consistency metric generalizes across scene categories, one might derive scene-type-specific ranking thresholds without retraining the full pipeline.
  • A natural next measurement would be the method's behavior on dynamic sequences where view consistency must also hold over time.

Load-bearing premise

The ranking step correctly identifies which primitives are truly redundant on the basis of geometric consistency across views.

What would settle it

On a test scene, renderings produced after merging according to the multi-view ranking show measurable increases in perceptual error or geometric distortion compared with the uncompressed 3DGS baseline.

Figures

Figures reproduced from arXiv: 2605.19304 by Beizhen Zhao, Dongxu Shen, Hao Wang, Sicheng Yu, Ziran Yin.

Figure 1
Figure 1. Figure 1: The pipeline of MMGS. First, we propose a multi-view 3D Gaussian ranking mechanism that resolves the geometric ambiguity of 2D gradients by validating candidate Gaussians at consistent 3D locations. Then, we design a global optimal transport algorithm to aggregate primitives into representative Gaussians based on ranking scores. Finally, we introduce an OT-based densification strategy that preserves the pa… view at source ↗
Figure 2
Figure 2. Figure 2: Comparison Results on Waymo [37] dataset. Visual differences are highlighted with red insets for better clarity. Our approach consistently outperforms other models on different scenes, demonstrating advantages in challenging scenarios. Best viewed in color [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Comparison Results on Mip-NeRF360 [34], Tanks & Temples [35] Deep Blending [36] dataset. Visual differences are highlighted with red insets for better clarity. Our approach consistently outperforms other models on different scenes, demonstrating advantages in challenging scenarios. Best viewed in color [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Ablation study on D. The influence of different KD-Tree depth on PSNR and the runtime of the aggregation algorithm. By choosing an appropriate M, we reduce the complexity by a factor of M compared to the global approach. In our implementation, with N ≈ 106 and M ≈ 1024, this linear scaling allows the entire reduction process to complete within a few seconds. Finally, the reconstruction of covariance matric… view at source ↗
Figure 5
Figure 5. Figure 5: Comparison Results. Visual differences are highlighted with red insets for better clarity. Our approach consistently outperforms other models on different scenes, demonstrating advantages in challenging scenarios. Best viewed in color. H Reproducibility Statement We have made every effort to ensure that the results presented in this paper are reproducible. All code and datasets will be made publicly availa… view at source ↗
Figure 6
Figure 6. Figure 6: Comparison Results. Visual differences are highlighted with red insets for better clarity. Our approach consistently outperforms other models on different scenes, demonstrating advantages in challenging scenarios. Best viewed in color. 19 [PITH_FULL_IMAGE:figures/full_fig_p019_6.png] view at source ↗
read the original abstract

While 3D Gaussian Splatting (3DGS) has revolutionized 3D reconstruction, it suffers from significant overhead due to massive redundant primitives. Existing compression methods typically rely on local sampling or fixed pruning thresholds, which often struggle to balance redundancy reduction with high-fidelity rendering. To address this, we propose a novel framework that formulates Gaussian optimization as a global geometric distribution matching problem. Specifically, our approach integrates three components: (1) we introduce a multi-view 3D Gaussian contribution ranking mechanism that filters primitives using geometric consistency instead of local heuristics; (2) we propose a global Optimal Transport (OT)-based aggregation algorithm that merges redundant primitives while preserving the underlying geometry; and (3) we design an OT-based densification operator that maintains the Gaussian's distributional properties for stable optimization. Our approach achieves state-of-the-art rendering quality with only \textbf{10$\%$} primitives and \textbf{10$\times$} accelerated training speeds compared to vanilla 3DGS.

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

1 major / 2 minor

Summary. The manuscript proposes MMGS, a compression framework for 3D Gaussian Splatting (3DGS). It formulates the problem as global geometric distribution matching via three components: (1) a multi-view 3D Gaussian contribution ranking mechanism based on geometric consistency rather than local heuristics, (2) a global Optimal Transport (OT)-based aggregation algorithm to merge redundant primitives while preserving scene geometry, and (3) an OT-based densification operator to maintain distributional properties during optimization. The central empirical claim is state-of-the-art rendering quality using only 10% of the primitives together with 10× accelerated training relative to vanilla 3DGS.

Significance. If the quantitative results hold, the work would be a meaningful contribution to efficient neural rendering by replacing heuristic pruning with a principled optimal-transport formulation of redundancy removal. The multi-view geometric-consistency ranking and the explicit preservation of distributional properties under merging are technically coherent extensions of existing OT ideas to the 3DGS setting and could influence subsequent compression research.

major comments (1)
  1. Abstract: the central claim of SOTA rendering quality with only 10% primitives and 10× training speedup is stated without any quantitative metrics, baselines, error bars, dataset specifications, or comparison tables. This absence prevents verification of the load-bearing assertion that the multi-view ranking plus OT aggregation preserves geometry while achieving the reported compression and speed gains.
minor comments (2)
  1. Notation: define OT and 3DGS at first use and ensure consistent symbol usage for the ranking score and transport cost throughout the method description.
  2. Presentation: the abstract would benefit from a single sentence summarizing the key experimental datasets and the magnitude of the quality metrics (e.g., PSNR or SSIM deltas) that support the SOTA claim.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback. We address the concern about the abstract below and will revise the manuscript to strengthen the presentation of our empirical claims.

read point-by-point responses

  1. Referee: Abstract: the central claim of SOTA rendering quality with only 10% primitives and 10× training speedup is stated without any quantitative metrics, baselines, error bars, dataset specifications, or comparison tables. This absence prevents verification of the load-bearing assertion that the multi-view ranking plus OT aggregation preserves geometry while achieving the reported compression and speed gains.

    Authors: We agree that the abstract would be strengthened by including representative quantitative results. In the revised manuscript we will update the abstract to report key metrics such as average PSNR on Mip-NeRF 360 and Tanks & Temples, the exact compression ratio achieved, training-time speedup relative to vanilla 3DGS, and a brief mention of the primary baselines. The full experimental section already contains the complete tables with error bars, per-scene breakdowns, and dataset specifications; the abstract revision will simply surface the headline numbers for immediate verification while preserving conciseness. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation is self-contained

full rationale

The paper formulates compression as a global geometric distribution matching problem using a multi-view ranking mechanism for redundancy detection followed by OT-based aggregation and densification. No equations, fitted parameters, or self-citations are shown that reduce the central claims (10% primitives with maintained quality) to inputs by construction. The ranking and OT steps are presented as direct applications of standard concepts to a new signal, with performance claims resting on empirical results rather than self-referential definitions or load-bearing prior work by the authors. This is the most common honest finding for a method paper that does not invoke uniqueness theorems or rename known patterns.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review yields no explicit free parameters, axioms, or invented entities; the method description relies on standard optimal transport and geometric consistency assumptions that are not detailed here.

pith-pipeline@v0.9.0 · 5719 in / 1171 out tokens · 37024 ms · 2026-05-20T07:09:18.396559+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

  • IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    we propose a global Optimal Transport (OT)-based aggregation algorithm that merges redundant primitives while preserving the underlying geometry... d²_W(Si,Sj)=Tr(Σi+Σj−2(Σi^{1/2}ΣjΣi^{1/2})^{1/2})+‖μi−μj‖₂²

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

40 extracted references · 40 canonical work pages

  1. [1]

    3d gaussian splatting as new era: A survey.IEEE Transactions on Visualization and Computer Graphics, 2024

    Ben Fei, Jingyi Xu, Rui Zhang, Qingyuan Zhou, Weidong Yang, and Ying He. 3d gaussian splatting as new era: A survey.IEEE Transactions on Visualization and Computer Graphics, 2024

    work page 2024

  2. [2]

    Recent advances in 3d gaussian splatting.Computational Visual Media, 10(4):613–642, 2024

    Tong Wu, Yu-Jie Yuan, Ling-Xiao Zhang, Jie Yang, Yan-Pei Cao, Ling-Qi Yan, and Lin Gao. Recent advances in 3d gaussian splatting.Computational Visual Media, 10(4):613–642, 2024

    work page 2024

  3. [3]

    3d gaussian splatting: Survey, technologies, challenges, and opportunities.IEEE Transactions on Circuits and Systems for Video Technology, 2025

    Yanqi Bao, Tianyu Ding, Jing Huo, Yaoli Liu, Yuxin Li, Wenbin Li, Yang Gao, and Jiebo Luo. 3d gaussian splatting: Survey, technologies, challenges, and opportunities.IEEE Transactions on Circuits and Systems for Video Technology, 2025

    work page 2025

  4. [4]

    Octree-gs: Towards consistent real-time rendering with lod-structured 3d gaussians.arXiv preprint arXiv:2403.17898, 2024

    Kerui Ren, Lihan Jiang, Tao Lu, Mulin Yu, Linning Xu, Zhangkai Ni, and Bo Dai. Octree- gs: Towards consistent real-time rendering with lod-structured 3d gaussians.arXiv preprint arXiv:2403.17898, 2024

    work page arXiv 2024

  5. [5]

    Scaffold- gs: Structured 3d gaussians for view-adaptive rendering

    Tao Lu, Mulin Yu, Linning Xu, Yuanbo Xiangli, Limin Wang, Dahua Lin, and Bo Dai. Scaffold- gs: Structured 3d gaussians for view-adaptive rendering. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 20654–20664, 2024

    work page 2024

  6. [6]

    Milena T Bagdasarian, Paul Knoll, Y Li, Florian Barthel, Anna Hilsmann, Peter Eisert, and Wieland Morgenstern. 3dgs. zip: A survey on 3d gaussian splatting compression methods. In Computer Graphics Forum, volume 44, page e70078. Wiley Online Library, 2025

    work page 2025

  7. [7]

    Atomgs: Atomizing gaussian splatting for high-fidelity radiance field.arXiv preprint arXiv:2405.12369, 2024

    Rong Liu, Rui Xu, Yue Hu, Meida Chen, and Andrew Feng. Atomgs: Atomizing gaussian splatting for high-fidelity radiance field.arXiv preprint arXiv:2405.12369, 2024

    work page arXiv 2024

  8. [8]

    Trimming the fat: Efficient compression of 3d gaussian splats through pruning.arXiv preprint arXiv:2406.18214, 2024

    Muhammad Salman Ali, Maryam Qamar, Sung-Ho Bae, and Enzo Tartaglione. Trimming the fat: Efficient compression of 3d gaussian splats through pruning.arXiv preprint arXiv:2406.18214, 2024

    work page arXiv 2024

  9. [9]

    Hac: Hash-grid assisted context for 3d gaussian splatting compression

    Yihang Chen, Qianyi Wu, Weiyao Lin, Mehrtash Harandi, and Jianfei Cai. Hac: Hash-grid assisted context for 3d gaussian splatting compression. InEuropean Conference on Computer Vision, pages 422–438. Springer, 2024

    work page 2024

  10. [10]

    Taming 3dgs: High-quality radiance fields with limited resources

    Saswat Subhajyoti Mallick, Rahul Goel, Bernhard Kerbl, Markus Steinberger, Francisco Vicente Carrasco, and Fernando De La Torre. Taming 3dgs: High-quality radiance fields with limited resources. InSIGGRAPH Asia 2024 Conference Papers, pages 1–11, 2024

    work page 2024

  11. [11]

    arXiv preprint arXiv:2511.04283 , year=

    Shiwei Ren, Tianci Wen, Yongchun Fang, and Biao Lu. Fastgs: Training 3d gaussian splatting in 100 seconds.arXiv preprint arXiv:2511.04283, 2025

    work page arXiv 2025

  12. [12]

    Mini-splatting: Representing scenes with a constrained number of gaussians

    Guangchi Fang and Bing Wang. Mini-splatting: Representing scenes with a constrained number of gaussians. InEuropean Conference on Computer Vision, pages 165–181. Springer, 2024

    work page 2024

  13. [13]

    Speedy- splat: Fast 3d gaussian splatting with sparse pixels and sparse primitives

    Alex Hanson, Allen Tu, Geng Lin, Vasu Singla, Matthias Zwicker, and Tom Goldstein. Speedy- splat: Fast 3d gaussian splatting with sparse pixels and sparse primitives. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 21537–21546, 2025

    work page 2025

  14. [14]

    Springer, 2008

    Cédric Villani et al.Optimal transport: old and new, volume 338. Springer, 2008

    work page 2008

  15. [15]

    Introduction to optimal transport.Notes of Course at University of Cambridge, 3, 2018

    Matthew Thorpe. Introduction to optimal transport.Notes of Course at University of Cambridge, 3, 2018

    work page 2018

  16. [16]

    Recent advances in optimal transport for machine learning.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024

    Eduardo Fernandes Montesuma, Fred Maurice Ngole Mboula, and Antoine Souloumiac. Recent advances in optimal transport for machine learning.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024. 10

    work page 2024

  17. [17]

    The expectation-maximization algorithm.IEEE Signal processing magazine, 13(6):47–60, 1996

    Todd K Moon. The expectation-maximization algorithm.IEEE Signal processing magazine, 13(6):47–60, 1996

    work page 1996

  18. [18]

    Dashgaussian: Optimizing 3d gaussian splatting in 200 seconds

    Youyu Chen, Junjun Jiang, Kui Jiang, Xiao Tang, Zhihao Li, Xianming Liu, and Yinyu Nie. Dashgaussian: Optimizing 3d gaussian splatting in 200 seconds. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 11146–11155, 2025

    work page 2025

  19. [19]

    Gaussianpro: 3d gaussian splatting with progressive propagation

    Kai Cheng, Xiaoxiao Long, Kaizhi Yang, Yao Yao, Wei Yin, Yuexin Ma, Wenping Wang, and Xuejin Chen. Gaussianpro: 3d gaussian splatting with progressive propagation. InForty-first International Conference on Machine Learning, 2024

    work page 2024

  20. [20]

    3d gaussian splatting as markov chain monte carlo.Advances in Neural Information Processing Systems, 37:80965–80986, 2024

    Shakiba Kheradmand, Daniel Rebain, Gopal Sharma, Weiwei Sun, Yang-Che Tseng, Hossam Isack, Abhishek Kar, Andrea Tagliasacchi, and Kwang Moo Yi. 3d gaussian splatting as markov chain monte carlo.Advances in Neural Information Processing Systems, 37:80965–80986, 2024

    work page 2024

  21. [21]

    Light- gaussian: Unbounded 3d gaussian compression with 15x reduction and 200+ fps.Advances in neural information processing systems, 37:140138–140158, 2024

    Zhiwen Fan, Kevin Wang, Kairun Wen, Zehao Zhu, Dejia Xu, Zhangyang Wang, et al. Light- gaussian: Unbounded 3d gaussian compression with 15x reduction and 200+ fps.Advances in neural information processing systems, 37:140138–140158, 2024

    work page 2024

  22. [22]

    Radsplat: Radiance field-informed gaussian splatting for robust real-time rendering with 900+ fps

    Michael Niemeyer, Fabian Manhardt, Marie-Julie Rakotosaona, Michael Oechsle, Daniel Duckworth, Rama Gosula, Keisuke Tateno, John Bates, Dominik Kaeser, and Federico Tombari. Radsplat: Radiance field-informed gaussian splatting for robust real-time rendering with 900+ fps. In2025 International Conference on 3D Vision (3DV), pages 134–144. IEEE, 2025

    work page 2025

  23. [23]

    Pup 3d-gs: Principled uncertainty pruning for 3d gaussian splatting

    Alex Hanson, Allen Tu, Vasu Singla, Mayuka Jayawardhana, Matthias Zwicker, and Tom Goldstein. Pup 3d-gs: Principled uncertainty pruning for 3d gaussian splatting. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 5949–5958, 2025

    work page 2025

  24. [24]

    Compact 3d gaussian splatting for static and dynamic radiance fields.arXiv preprint arXiv:2408.03822, 2024

    Joo Chan Lee, Daniel Rho, Xiangyu Sun, Jong Hwan Ko, and Eunbyung Park. Compact 3d gaussian splatting for static and dynamic radiance fields.arXiv preprint arXiv:2408.03822, 2024

    work page arXiv 2024

  25. [25]

    Lp-3dgs: Learning to prune 3d gaussian splatting.Advances in Neural Information Processing Systems, 37:122434–122457, 2024

    Zhaoliang Zhang, Tianchen Song, Yongjae Lee, Li Yang, Cheng Peng, Rama Chellappa, and Deliang Fan. Lp-3dgs: Learning to prune 3d gaussian splatting.Advances in Neural Information Processing Systems, 37:122434–122457, 2024

    work page 2024

  26. [26]

    Stopthepop: Sorted gaussian splatting for view-consistent real-time rendering.ACM Transactions on Graphics (TOG), 43(4):1–17, 2024

    Lukas Radl, Michael Steiner, Mathias Parger, Alexander Weinrauch, Bernhard Kerbl, and Markus Steinberger. Stopthepop: Sorted gaussian splatting for view-consistent real-time rendering.ACM Transactions on Graphics (TOG), 43(4):1–17, 2024

    work page 2024

  27. [27]

    Flashgs: Efficient 3d gaussian splatting for large-scale and high-resolution rendering

    Guofeng Feng, Siyan Chen, Rong Fu, Zimu Liao, Yi Wang, Tao Liu, Boni Hu, Linning Xu, Zhilin Pei, Hengjie Li, et al. Flashgs: Efficient 3d gaussian splatting for large-scale and high-resolution rendering. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 26652–26662, 2025

    work page 2025

  28. [28]

    3dgs-lm: Faster gaussian- splatting optimization with levenberg-marquardt

    Lukas Höllein, Aljaž Božiˇc, Michael Zollhöfer, and Matthias Nießner. 3dgs-lm: Faster gaussian- splatting optimization with levenberg-marquardt. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 26740–26750, 2025

    work page 2025

  29. [29]

    3dgs2: Near second-order converging 3d gaussian splatting

    Lei Lan, Tianjia Shao, Zixuan Lu, Yu Zhang, Chenfanfu Jiang, and Yin Yang. 3dgs2: Near second-order converging 3d gaussian splatting. InProceedings of the Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference Papers, pages 1–10, 2025

    work page 2025

  30. [30]

    Gaussian herding across pens: An optimal transport perspective on global gaussian reduction for 3dgs.arXiv preprint arXiv:2506.09534, 2025

    Tao Wang, Mengyu Li, Geduo Zeng, Cheng Meng, and Qiong Zhang. Gaussian herding across pens: An optimal transport perspective on global gaussian reduction for 3dgs.arXiv preprint arXiv:2506.09534, 2025

    work page arXiv 2025

  31. [31]

    The wasserstein distance and approximation theorems.Probability Theory and Related Fields, 70(1):117–129, 1985

    Ludger Rüschendorf. The wasserstein distance and approximation theorems.Probability Theory and Related Fields, 70(1):117–129, 1985

    work page 1985

  32. [32]

    On the bures–wasserstein distance between positive definite matrices.Expositiones mathematicae, 37(2):165–191, 2019

    Rajendra Bhatia, Tanvi Jain, and Yongdo Lim. On the bures–wasserstein distance between positive definite matrices.Expositiones mathematicae, 37(2):165–191, 2019. 11

    work page 2019

  33. [33]

    On a formula for the l2 wasserstein metric between measures on euclidean and hilbert spaces.Mathematische Nachrichten, 147(1):185–203, 1990

    Matthias Gelbrich. On a formula for the l2 wasserstein metric between measures on euclidean and hilbert spaces.Mathematische Nachrichten, 147(1):185–203, 1990

    work page 1990

  34. [34]

    Mip- nerf 360: Unbounded anti-aliased neural radiance fields

    Jonathan T Barron, Ben Mildenhall, Dor Verbin, Pratul P Srinivasan, and Peter Hedman. Mip- nerf 360: Unbounded anti-aliased neural radiance fields. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5470–5479, 2022

    work page 2022

  35. [35]

    Tanks and temples: Bench- marking large-scale scene reconstruction.ACM Transactions on Graphics, 36(4), 2017

    Arno Knapitsch, Jaesik Park, Qian-Yi Zhou, and Vladlen Koltun. Tanks and temples: Bench- marking large-scale scene reconstruction.ACM Transactions on Graphics, 36(4), 2017

    work page 2017

  36. [36]

    Deep blending for free-viewpoint image-based rendering.ACM Transactions on Graphics (ToG), 37(6):1–15, 2018

    Peter Hedman, Julien Philip, True Price, Jan-Michael Frahm, George Drettakis, and Gabriel Brostow. Deep blending for free-viewpoint image-based rendering.ACM Transactions on Graphics (ToG), 37(6):1–15, 2018

    work page 2018

  37. [37]

    Scalability in perception for autonomous driving: Waymo open dataset

    Pei Sun, Henrik Kretzschmar, Xerxes Dotiwalla, Aurelien Chouard, Vijaysai Patnaik, Paul Tsui, James Guo, Yin Zhou, Yuning Chai, Benjamin Caine, Vijay Vasudevan, Wei Han, Jiquan Ngiam, Hang Zhao, Aleksei Timofeev, Scott Ettinger, Maxim Krivokon, Amy Gao, Aditya Joshi, Yu Zhang, Jonathon Shlens, Zhifeng Chen, and Dragomir Anguelov. Scalability in perception...

    work page 2020

  38. [38]

    Image quality assessment: from error visibility to structural similarity.IEEE transactions on image processing, 13(4):600– 612, 2004

    Zhou Wang, Alan C Bovik, Hamid R Sheikh, and Eero P Simoncelli. Image quality assessment: from error visibility to structural similarity.IEEE transactions on image processing, 13(4):600– 612, 2004

    work page 2004

  39. [39]

    The unrea- sonable effectiveness of deep features as a perceptual metric

    Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shechtman, and Oliver Wang. The unrea- sonable effectiveness of deep features as a perceptual metric. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 586–595, 2018

    work page 2018

  40. [40]

    3d gaussian splatting for real-time radiance field rendering.ACM Trans

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimkühler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering.ACM Trans. Graph., 42(4):139–1, 2023. A LLM Declaration and Impact Statements The core method development in this research does not involve LLMs as any important, original, or non-standard components. This paper presents ...

    work page 2023