Pith. sign in

REVIEW 3 major objections 5 minor 75 references

Point Cloud Unsupervised Pre-training via 3D Gaussian Splatting

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The paper claims that rendering with 3D Gaussian Splatting, instead of volume rendering, can pre-train point cloud encoders about 9x faster and with less than a quarter of the memory while still transferring to detection, segmentation…

desk verdict New GS-based rendering pretext for point cloud SSL with broad transfer results, but the 9x speedup and <0.25x memory claims rest on shaky extrapolation and need direct measurement. read the letter →

arxiv 2411.18667 v1 pith:GMIYSCS5 submitted 2024-11-27 cs.CV

Hao Liu , Minglin Chen , Yanni Ma , Haihong Xiao , Ying He This is my paper · ORCID
classification cs.CV
keywords pointcloudself-supervisedlearning3DGaussianSplattingneuralrenderingpre-trainingmaskedmodelingobjectdetectionsemanticsegmentationinstancescenereconstruction
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

This paper tries to establish that a self-supervised point cloud encoder can be pre-trained by using a fast 3D Gaussian Splatting renderer as the pretext task, replacing the expensive volume rendering used by prior work. The proposed framework, GS3, back-projects RGB-D images into colored point clouds, encodes them with a point cloud network, predicts per-point 3D Gaussians, and renders RGB images that are compared with the real input images. The pre-trained encoder is then fine-tuned for 3D detection, semantic and instance segmentation, and scene reconstruction. If correct, rendering-based self-supervised learning for point clouds becomes practical on a single GPU, with about a 9x pre-training speedup and under 0.25x the memory of the previous volume-rendering baseline, while matching or slightly exceeding its downstream accuracy.

What carries the argument

The machinery is point-aligned Gaussian prediction plus a differentiable tile-based rasterizer: a feed-forward network reads the encoded point features and predicts, for each point, Gaussian center offsets, covariances, opacities, and colors, so that each point anchors one or more 3D Gaussians in a point-aligned manner. The rasterizer projects all Gaussians to the image plane, composites them with alpha blending into an RGB image, and the image is compared with the input through an MSE color loss and an LPIPS perceptual loss. Masked point modeling (MPM) completes the design by masking 50% of the points and rendering the scene from the visible ones, forcing the encoder to capture contextual geometry and appearance.

What would settle it

Measure the volume-rendering baseline's per-batch time and GPU memory at 76,800 sampling rays on a single A100 with the same batch size and resolution, then compare with GS3's reported 2.67 s/batch and 10.3 GB/batch. If the baseline's measured runtime is not roughly 9x larger or its memory is not more than 4x larger, the headline efficiency ratio is an artifact of extrapolation.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the per-ray sampling cost of volume rendering is not needed for point cloud self-supervised learning: a generalizable 3D Gaussian Splatting renderer, fed with Gaussians predicted point-aligned from encoder features, can produce the RGB reconstructions that drive representation learning. Because the tile-based rasterizer is much cheaper, the same two-view RGB-D setup becomes roughly 9x faster to pre-train and uses under 0.25x the memory. The rendered-versus-real image comparison forces the encoder to retain the geometric and appearance information needed to place and color thousands of Gaussians, and the resulting features transfer to four downstream task families, with consistent gains over training-from-scratch baselines and accuracy comparable to the prior rendering-based approach.

Load-bearing premise

The load-bearing premise is that the volume-rendering baseline's runtime scales linearly from 4,800 to 76,800 sampling rays while its memory stays at the 4,800-ray level, so the claimed 9x speedup and sub-0.25x memory use reflect matched settings rather than extrapolation.

Editorial extensions

If this is right

  • Pre-training with GS3 on ScanNet lifts VoteNet's mAP@0.5 by 3.0 points on SUN RGB-D, matching or exceeding the gains of the Ponder baseline.
  • On S3DIS Area-5, a GS3-pretrained MinkUNet at 2 cm voxels reaches 70.1 mIoU, 1.6 points above the same architecture trained from scratch.
  • The same encoder initialization improves PointGroup instance segmentation and ConvONet scene reconstruction, with volumetric IoU reaching 79.7% on the Synthetic Indoor Scene dataset.
  • The claimed pre-training cost is 2.67 s/batch and 10.3 GB/batch, versus an estimated 23.36 s/batch for the volume-rendering baseline at 76,800 sampling rays.
  • Masking 50% of points during pre-training is the best setting, but every mask ratio from 0% to 90% still improves the downstream baseline.
  • Adding a third input view raises segmentation mIoU to 70.5%, indicating that the framework can absorb more views when pre-training budget allows.

Reading between the lines

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

  • The paper does not test, but the three-view ablation suggests the framework has headroom: more images or higher resolution may push downstream gains further with the same design.
  • Because supervision is only RGB and depth is used just to lift points into 3D, the framework could in principle be applied to point clouds obtained without depth sensors; whether the pretext still works with estimated depth is an open question.
  • A fair efficiency comparison would measure a volume-rendering baseline that has also been optimized for speed, rather than the original implementation, before treating the 9x speedup as intrinsic to Gaussian splatting.
  • The mask-ratio robustness observed in the ablations suggests the rendering pretext is forgiving of incomplete input, which could matter for pre-training on noisy or partially observed scans.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes GS3, a self-supervised pre-training framework for point cloud encoders. It back-projects sparse-view RGB-D images into colored point clouds, extracts point-wise features with PointNet++ or SR-UNet, predicts per-point 3D Gaussian parameters, and renders the scene with a tile-based rasterizer; the pre-training loss is an MSE color loss plus an LPIPS loss. The pre-trained encoders are fine-tuned on 3D object detection, semantic and instance segmentation, and scene reconstruction, where they improve over the corresponding baselines. The paper further claims roughly a 9x pre-training speedup and less than 0.25x memory usage relative to the rendering-based baseline Ponder.

Significance. If the results hold, the paper makes a useful practical contribution by replacing volume rendering with 3D Gaussian splatting for rendering-based point cloud SSL, enabling lower-cost pre-training on ScanNet and consistent downstream gains across four tasks. The experiments are broad and internally consistent, with ablations on mask ratio, supervision type, number of views, and resolution, and per-category results in the appendix. The central transferability claim is plausible and supported by improvements over strong baselines (e.g., VoteNet +3.0 mAP@0.5, MinkUNet +1.5-1.6 mIoU). The main unresolved issue is the efficiency claim, which rests on an extrapolated and unmatched baseline comparison; this is fixable with additional measurement or with a revised claim.

major comments (3)
  1. [§4.2.1, Table 2 and Fig. 1] The headline efficiency claims are not established. The Ponder row at 76,800 rays (23.36 s/batch) is an estimate obtained by linearly scaling the measured 4,800-ray result (1.46 s/batch), and the manuscript provides no evidence that Ponder's per-batch runtime scales linearly with the number of sampling rays. In addition, Ponder's memory is reported only at 4,800 rays, so the '<0.25x memory' claim in the abstract and Fig. 1 is not directly measured; the only directly measured ratio is 10.3/38.4 ≈ 0.27, which does not support '<0.25x'. Please measure Ponder at the same ray count and image resolution, or revise the abstract, Fig. 1, and conclusion to state the actual measured comparison.
  2. [Table 2] The workloads being compared are not shown to be matched. GS3 renders all 76,800 pixels of a 320×240 image, while Ponder is reported at 4,800 sampling rays; the caption does not state Ponder's image resolution, sampling pattern, batch size, or number of views. If Ponder is run under different settings, the per-batch time and memory numbers are not commensurate. Please report the same workload parameters for both methods, including batch size and pre-training epochs.
  3. [Tables 1, 4, and 6] The differences between GS3 and Ponder on common downstream tasks are very small (36.7 vs 36.6 mAP@0.5 on SUN RGB-D; 73.4 vs 73.5 mIoU on ScanNet v2; 79.7 vs 80.2 IoU on Synthetic Indoor Scene), but no variance or repeated-seed information is given. Without such information, the claim that GS3 is comparable to Ponder is not distinguishable from run-to-run noise. Please add repeated-run statistics or explicitly state that these are single runs, and adjust the wording accordingly.
minor comments (5)
  1. [Section 3.2] SR-UNet is cited as [57], but SR-UNet (MinkUNet) is introduced in [8]/[9]; please correct the citation.
  2. [Appendix A.1] 'PoinbtNet++' is a typo and should be 'PointNet++'.
  3. [Table 6 caption] 'improment' should be 'improvement'.
  4. [Section 4.2.1, text after Table 3] 'H3Net' should be 'H3DNet'.
  5. [Abstract] The grammar of 'we back-project ... and uses a tile-based rasterizer' should be made consistent ('use') for readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: downstream transfer is measured on held-out benchmarks, pre-training constants are fixed a priori, and the efficiency comparison is an extrapolation concern rather than a definitional reduction.

full rationale

This is an empirical method paper, not a derivation. The pre-training objective in Eq. 7 and Eq. 8 directly compares rendered images to real input images, and the point cloud encoder is then fine-tuned on held-out downstream benchmarks (SUN RGB-D, ScanNet v2, S3DIS, Synthetic Indoor Scene). None of the hyperparameters reported in Section 4.1.2, such as the LPIPS weight lambda = 0.05, the 50% mask ratio, or the two-view setting, is fitted to any downstream result. The claimed speedup and memory reduction in Table 2 and Figure 1 depend on an extrapolated Ponder runtime at 76,800 rays, which is a measurement-fairness concern, not a case where a fitted input is renamed as a prediction; the Ponder numbers are external baseline measurements, and the GS3 numbers are its own measured overheads. There are no self-citations carrying the central claim, no uniqueness theorem imported from prior work, and no known result merely renamed in new coordinates. The paper's core comparison, rendering-based pre-training versus alternative self-supervised methods, is therefore self-contained and not circular, although the efficiency claim should be re-verified with directly measured Ponder numbers.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The framework introduces no new physical or mathematical entities; it reuses 3D Gaussians, tile-based rendering, and standard SSL losses. The most fragile ledger item is the extrapolated baseline runtime for Ponder, which carries the efficiency claim.

free parameters (4)
  • LPIPS loss weight lambda = 0.05
    Set to pixelSplat's value (Section 3.4); not tuned in this paper, but directly weights one of the two training losses.
  • Mask ratio = 50%
    Selected via ablation on S3DIS segmentation (Table 7); 50% gives the best mIoU among 0%, 25%, 50%, 75%, 90%.
  • Number of input views = 2
    Chosen as an efficiency/performance tradeoff; ablation shows 3 views gives higher mIoU (70.5 vs 70.1) but more cost (Table 9).
  • Input image resolution = 320x240
    Chosen as a tradeoff; 512x384 gives a further +0.4 mIoU over the chosen setting (Table 10).
assumptions (4)
  • domain assumption A view-reconstruction pretext on the training views themselves, with 50% of points masked, forces the encoder to learn transferable geometry and appearance features.
    This is the core SSL premise behind the render-and-compare objective; validated only indirectly through downstream fine-tuning, not proven.
  • domain assumption Back-projected RGB-D colored point clouds contain sufficient geometric and color information to supervise scene Gaussians for the pretext task.
    Assumes depth values and camera poses from ScanNet are reliable enough; no sensitivity analysis to depth noise is provided.
  • standard math The tile-based 3D Gaussian rasterizer and the feed-forward Gaussian prediction heads are differentiable and pass gradients to the point cloud encoder.
    Relies on established differentiable rendering from 3DGS [25] and generalizable GS [3]; not re-derived here.
  • ad hoc to paper Linearly extrapolating Ponder's runtime per batch from 4,800 rays to 76,800 rays gives a valid comparison.
    This assumption underlies the headline 9x speedup; the paper itself states the 76,800-ray Ponder time is estimated (Table 2 caption).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Point Cloud Unsupervised Pre-training via 3D Gaussian Splatting." pith.science (2026). https://pith.science/paper/GMIYSCS5

@misc{pith2026241118667,
  author       = {Pith},
  title        = {Pith review of: Point Cloud Unsupervised Pre-training via 3D Gaussian Splatting},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GMIYSCS5}},
  note         = {Machine review of arXiv:2411.18667}
}
abstract

Pre-training on large-scale unlabeled datasets contribute to the model achieving powerful performance on 3D vision tasks, especially when annotations are limited. However, existing rendering-based self-supervised frameworks are computationally demanding and memory-intensive during pre-training due to the inherent nature of volume rendering. In this paper, we propose an efficient framework named GS$^3$ to learn point cloud representation, which seamlessly integrates fast 3D Gaussian Splatting into the rendering-based framework. The core idea behind our framework is to pre-train the point cloud encoder by comparing rendered RGB images with real RGB images, as only Gaussian points enriched with learned rich geometric and appearance information can produce high-quality renderings. Specifically, we back-project the input RGB-D images into 3D space and use a point cloud encoder to extract point-wise features. Then, we predict 3D Gaussian points of the scene from the learned point cloud features and uses a tile-based rasterizer for image rendering. Finally, the pre-trained point cloud encoder can be fine-tuned to adapt to various downstream 3D tasks, including high-level perception tasks such as 3D segmentation and detection, as well as low-level tasks such as 3D scene reconstruction. Extensive experiments on downstream tasks demonstrate the strong transferability of the pre-trained point cloud encoder and the effectiveness of our self-supervised learning framework. In addition, our GS$^3$ framework is highly efficient, achieving approximately 9$\times$ pre-training speedup and less than 0.25$\times$ memory cost compared to the previous rendering-based framework Ponder.

Figures

Figures reproduced from arXiv: 2411.18667 by the authors.

Figure 1
Figure 1. Comparison of 3D detection performance mAP@0.5, 3D segmentation accuracy mIoU, pre-training time and memory con￾sumption of Ponder [21] and our GS3 . The pre-training time and memory usage of our method are measured at a rendered image resolution of 320 × 240. Due to limited computational resources, the pre-training time of Ponder with 76,800 sampling rays is esti￾mated based on its result with 4,800 rays. Memory co… view at source ↗
Figure 2
Figure 2. The overall framework of the proposed GS [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The network architecture of our feature encoder. (a) SR-UNet and (b) PointNet++. For SR-UNet, each sparse (de)convolution [PITH_FULL_IMAGE:figures/full_fig_p013_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Qualitative results of our fine-tuned model on downstream (a) 3D semantic segmentation and (b) 3D instance segmentation tasks. [PITH_FULL_IMAGE:figures/full_fig_p014_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

75 extracted references · 64 canonical work pages

  1. [57]

    Pointcontrast: Unsupervised pre- training for 3D point cloud understanding

    Saining Xie, Jiatao Gu, Demi Guo, Charles R Qi, Leonidas Guibas, and Or Litany. Pointcontrast: Unsupervised pre- training for 3D point cloud understanding. In Proceedings of the European Conference on Computer Vision (ECCV) , pages 574–591. Springer, 2020. 2, 4, 6

    work page 2020

  2. [9]

    4D spatio-temporal convnets: Minkowski convolutional neural networks

    Christopher Choy, JunYoung Gwak, and Silvio Savarese. 4D spatio-temporal convnets: Minkowski convolutional neural networks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pages 3075–3084, 2019. 3

    work page 2019

  3. [1]

    3D semantic parsing of large-scale indoor spaces

    Iro Armeni, Ozan Sener, Amir R Zamir, Helen Jiang, Ioannis Brilakis, Martin Fischer, and Silvio Savarese. 3D semantic parsing of large-scale indoor spaces. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion (CVPR), pages 1534–1543, 2016. 6, 7

    work page 2016

  4. [2]

    Mip-nerf: A multiscale representation for anti-aliasing neu- ral radiance fields

    Jonathan T Barron, Ben Mildenhall, Matthew Tancik, Peter Hedman, Ricardo Martin-Brualla, and Pratul P Srinivasan. Mip-nerf: A multiscale representation for anti-aliasing neu- ral radiance fields. In Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision (ICCV), pages 5855– 5864, 2021. 3

    work page 2021

  5. [3]

    pixelsplat: 3D gaussian splats from im- age pairs for scalable generalizable 3D reconstruction

    David Charatan, Sizhe Lester Li, Andrea Tagliasacchi, and Vincent Sitzmann. pixelsplat: 3D gaussian splats from im- age pairs for scalable generalizable 3D reconstruction. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition (CVPR), pages 19457–19467,

    work page

  6. [4]

    Tensorf: Tensorial radiance fields

    Anpei Chen, Zexiang Xu, Andreas Geiger, Jingyi Yu, and Hao Su. Tensorf: Tensorial radiance fields. In Proceedings of the European Conference on Computer Vision (ECCV) , pages 333–350. Springer, 2022. 3

    work page 2022

  7. [5]

    Multi-view 3D object detection network for autonomous driving

    Xiaozhi Chen, Huimin Ma, Ji Wan, Bo Li, and Tian Xia. Multi-view 3D object detection network for autonomous driving. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pages 1907–1915, 2017. 3

    work page 1907

  8. [6]

    4Dcon- trast: Contrastive learning with dynamic correspondences for 3D scene understanding

    Yujin Chen, Matthias Nießner, and Angela Dai. 4Dcon- trast: Contrastive learning with dynamic correspondences for 3D scene understanding. In Proceedings of the Euro- pean Conference on Computer Vision (ECCV) , pages 543–

    work page

Show all 75 references
  1. [7]

    Mvsplat: Efficient 3D gaussian splatting from sparse multi-view images

    Yuedong Chen, Haofei Xu, Chuanxia Zheng, Bohan Zhuang, Marc Pollefeys, Andreas Geiger, Tat-Jen Cham, and Jianfei Cai. Mvsplat: Efficient 3D gaussian splatting from sparse multi-view images. In Proceedings of the European Confer- ence on Computer Vision (ECCV), 2024. 3, 4

    2024

  2. [8]

    4D spatio-temporal convnets: Minkowski convolutional neural networks

    Christopher Choy, JunYoung Gwak, and Silvio Savarese. 4D spatio-temporal convnets: Minkowski convolutional neural networks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pages 3075–3084, 2019. 6, 7, 13, 14

    2019

  3. [10]

    Spconv: Spatially sparse convolution library

    Spconv Contributors. Spconv: Spatially sparse convolution library. 2022. 3

    2022

  4. [11]

    Scannet: Richly-annotated 3D reconstructions of indoor scenes

    Angela Dai, Angel X Chang, Manolis Savva, Maciej Hal- ber, Thomas Funkhouser, and Matthias Nießner. Scannet: Richly-annotated 3D reconstructions of indoor scenes. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , pages 5828–5839, 2017. 1...

    2017

  5. [12]

    A point set generation network for 3d object reconstruction from a sin- gle image

    Haoqiang Fan, Hao Su, and Leonidas J Guibas. A point set generation network for 3d object reconstruction from a sin- gle image. In Proceedings of the IEEE Conference on Com- puter Vision and Pattern Recognition (CVPR) , pages 605– 613, 2017. 2

    2017

  6. [13]

    3D semantic segmentation with submanifold sparse convolutional networks

    Benjamin Graham, Martin Engelcke, and Laurens Van Der Maaten. 3D semantic segmentation with submanifold sparse convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 9224–9232, 2018. 7

    2018

  7. [14]

    3D semantic segmentation with submanifold sparse convolutional networks

    Benjamin Graham, Martin Engelcke, and Laurens Van Der Maaten. 3D semantic segmentation with submanifold sparse convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 9224–9232, 2018. 3

    2018

  8. [15]

    Multiple view ge- ometry in computer vision

    Richard Hartley and Andrew Zisserman. Multiple view ge- ometry in computer vision . Cambridge university press,

  9. [16]

    Masked autoencoders are scalable vision learners

    Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Doll´ar, and Ross Girshick. Masked autoencoders are scalable vision learners. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 16000–16009, 2022. 1

    2022

  10. [17]

    Dyco3d: Robust instance segmentation of 3D point clouds through dynamic convolution

    Tong He, Chunhua Shen, and Anton Van Den Hengel. Dyco3d: Robust instance segmentation of 3D point clouds through dynamic convolution. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 354–363, 2021. 7

    2021

  11. [18]

    Masked autoencoder for self-supervised pre-training on lidar point clouds

    Georg Hess, Johan Jaxing, Elias Svensson, David Hager- man, Christoffer Petersson, and Lennart Svensson. Masked autoencoder for self-supervised pre-training on lidar point clouds. In Proceedings of the IEEE/CVF winter conference on applications of computer vision, pages 350–35...

    2023

  12. [19]

    3D-SIS: 3D se- mantic instance segmentation of rgb-d scans

    Ji Hou, Angela Dai, and Matthias Nießner. 3D-SIS: 3D se- mantic instance segmentation of rgb-d scans. InProceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition (CVPR), pages 4421–4430, 2019. 7

    2019

  13. [20]

    Exploring data-efficient 3D scene understanding with contrastive scene contexts

    Ji Hou, Benjamin Graham, Matthias Nießner, and Saining Xie. Exploring data-efficient 3D scene understanding with contrastive scene contexts. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 15587–15597, 2021. 2

    2021

  14. [21]

    Ponder: Point cloud pre-training via neural rendering

    Di Huang, Sida Peng, Tong He, Honghui Yang, Xiaowei Zhou, and Wanli Ouyang. Ponder: Point cloud pre-training via neural rendering. In Proceedings of the IEEE/CVF In- ternational Conference on Computer Vision (ICCV) , pages 16089–16098, 2023. 1, 2, 5, 6, 7

    2023

  15. [22]

    Spatio-temporal self-supervised representation learning for 3D point clouds

    Siyuan Huang, Yichen Xie, Song-Chun Zhu, and Yixin Zhu. Spatio-temporal self-supervised representation learning for 3D point clouds. In Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision (ICCV), pages 6535– 6545, 2021. 2, 6

    2021

  16. [23]

    Pointgroup: Dual-set point group- ing for 3D instance segmentation

    Li Jiang, Hengshuang Zhao, Shaoshuai Shi, Shu Liu, Chi- Wing Fu, and Jiaya Jia. Pointgroup: Dual-set point group- ing for 3D instance segmentation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 4867–4876, 2020. 7, 14

    2020

  17. [24]

    Guided point contrastive learn- 9 ing for semi-supervised point cloud semantic segmentation

    Li Jiang, Shaoshuai Shi, Zhuotao Tian, Xin Lai, Shu Liu, Chi-Wing Fu, and Jiaya Jia. Guided point contrastive learn- 9 ing for semi-supervised point cloud semantic segmentation. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pages 6423–6432, 2021. 2

    2021

  18. [25]

    3D gaussian splatting for real-time radiance field rendering

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk ¨uhler, and George Drettakis. 3D gaussian splatting for real-time radiance field rendering. ACM Transactions on Graphics (ToG), 42(4):1–14, 2023. 3

    2023

  19. [26]

    PointPillars: Fast encoders for object detection from point clouds

    Alex H Lang, Sourabh V ora, Holger Caesar, Lubing Zhou, Jiong Yang, and Oscar Beijbom. PointPillars: Fast encoders for object detection from point clouds. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 12697–12705, 2019. 3

    2019

  20. [27]

    Compact 3D gaussian representation for radiance field

    Joo Chan Lee, Daniel Rho, Xiangyu Sun, Jong Hwan Ko, and Eunbyung Park. Compact 3D gaussian representation for radiance field. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 21719–21728, 2024. 3

    2024

  21. [28]

    Deep continuous fusion for multi-sensor 3D object detection

    Ming Liang, Bin Yang, Shenlong Wang, and Raquel Urtasun. Deep continuous fusion for multi-sensor 3D object detection. In Proceedings of the European Conference on Computer Vi- sion (ECCV), 2018. 3

    2018

  22. [29]

    Semantic context encoding for accurate 3D point cloud segmentation

    Hao Liu, Yulan Guo, Yanni Ma, Yinjie Lei, and Gongjian Wen. Semantic context encoding for accurate 3D point cloud segmentation. IEEE Transactions on Multimedia (TMM), 23: 2045–2055, 2021. 3

    2021

  23. [30]

    Masked discrimi- nation for self-supervised learning on point clouds

    Haotian Liu, Mu Cai, and Yong Jae Lee. Masked discrimi- nation for self-supervised learning on point clouds. In Pro- ceedings of the European Conference on Computer Vision (ECCV), pages 657–675. Springer, 2022. 2, 4, 6

    2022

  24. [31]

    Cen- tertube: Tracking multiple 3D objects with 4d tubelets in dynamic point clouds

    Hao Liu, Yanni Ma, Qingyong Hu, and Yulan Guo. Cen- tertube: Tracking multiple 3D objects with 4d tubelets in dynamic point clouds. IEEE Transactions on Multimedia (TMM), 2023. 3

    2023

  25. [32]

    Anchor- point: Query design for transformer-based 3D object detec- tion and tracking

    Hao Liu, Yanni Ma, Hanyun Wang, and Yulan Guo. Anchor- point: Query design for transformer-based 3D object detec- tion and tracking. IEEE Transactions on Intelligent Trans- portation Systems (TITS), 2023. 3

    2023

  26. [33]

    Neuraludf: Learning unsigned distance fields for multi-view reconstruction of surfaces with arbitrary topolo- gies

    Xiaoxiao Long, Cheng Lin, Lingjie Liu, Yuan Liu, Peng Wang, Christian Theobalt, Taku Komura, and Wenping Wang. Neuraludf: Learning unsigned distance fields for multi-view reconstruction of surfaces with arbitrary topolo- gies. In Proceedings of the IEEE/CVF Conference on Com- ...

    2023

  27. [34]

    Decoupled weight decay regularization

    I Loshchilov. Decoupled weight decay regularization. arXiv preprint arXiv:1711.05101, 2017. 5

    2017 arXiv

  28. [35]

    Sgdr: Stochas- tic gradient descent with warm restarts

    Ilya Loshchilov and Frank Hutter. Sgdr: Stochas- tic gradient descent with warm restarts. arXiv preprint arXiv:1608.03983, 2016. 5

    2016 arXiv

  29. [36]

    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. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 20654–20664, 2024. 3

    2024

  30. [37]

    Srinivasan, Matthew Tancik, Jonathan T

    Ben Mildenhall, Pratul P. Srinivasan, Matthew Tancik, Jonathan T. Barron, Ravi Ramamoorthi, and Ren Ng. Nerf: Representing scenes as neural radiance fields for view syn- thesis. In Proceedings of the European Conference on Com- puter Vision (ECCV), 2020. 3, 6

    2020

  31. [38]

    Occupancy-mae: Self-supervised pre-training large- scale lidar point clouds with masked occupancy autoen- coders

    Chen Min, Liang Xiao, Dawei Zhao, Yiming Nie, and Bin Dai. Occupancy-mae: Self-supervised pre-training large- scale lidar point clouds with masked occupancy autoen- coders. IEEE Transactions on Intelligent Vehicles (TIV) ,

  32. [39]

    An end-to- end transformer model for 3D object detection

    Ishan Misra, Rohit Girdhar, and Armand Joulin. An end-to- end transformer model for 3D object detection. In Proceed- ings of the IEEE/CVF International Conference on Com- puter Vision (ICCV), pages 2906–2917, 2021. 6

    2021

  33. [40]

    Instant neural graphics primitives with a mul- tiresolution hash encoding

    Thomas M ¨uller, Alex Evans, Christoph Schied, and Alexan- der Keller. Instant neural graphics primitives with a mul- tiresolution hash encoding. ACM transactions on graphics (TOG), 41(4):1–15, 2022. 3

    2022

  34. [41]

    Masked autoencoders for point cloud self-supervised learning

    Yatian Pang, Wenxiao Wang, Francis EH Tay, Wei Liu, Yonghong Tian, and Li Yuan. Masked autoencoders for point cloud self-supervised learning. In Proceedings of the Euro- pean Conference on Computer Vision (ECCV) , pages 604–

  35. [42]

    Deepsdf: Learning con- tinuous signed distance functions for shape representation

    Jeong Joon Park, Peter Florence, Julian Straub, Richard Newcombe, and Steven Lovegrove. Deepsdf: Learning con- tinuous signed distance functions for shape representation. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition (CVPR), pages 165–1...

    2019

  36. [43]

    Convolutional occupancy networks

    Songyou Peng, Michael Niemeyer, Lars Mescheder, Marc Pollefeys, and Andreas Geiger. Convolutional occupancy networks. In Proceedings of the European Conference on Computer Vision (ECCV), pages 523–540. Springer, 2020. 7

    2020

  37. [44]

    PointNet: Deep learning on point sets for 3D classification and segmentation

    Charles R Qi, Hao Su, Kaichun Mo, and Leonidas J Guibas. PointNet: Deep learning on point sets for 3D classification and segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 652–660, 2017. 7, 13

    2017

  38. [45]

    Pointnet++: Deep hierarchical feature learning on point sets in a metric space

    Charles Ruizhongtai Qi, Li Yi, Hao Su, and Leonidas J Guibas. Pointnet++: Deep hierarchical feature learning on point sets in a metric space. Proceedings of the Advances in Neural Information Processing Systems (NeurIPS), 30, 2017. 3, 4, 7

    2017

  39. [46]

    Deep hough voting for 3D object detection in point clouds

    Charles R Qi, Or Litany, Kaiming He, and Leonidas J Guibas. Deep hough voting for 3D object detection in point clouds. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pages 9277–9286,

  40. [47]

    Randomrooms: Unsupervised pre- training from synthetic shapes and randomized layouts for 3D object detection

    Yongming Rao, Benlin Liu, Yi Wei, Jiwen Lu, Cho-Jui Hsieh, and Jie Zhou. Randomrooms: Unsupervised pre- training from synthetic shapes and randomized layouts for 3D object detection. In Proceedings of the IEEE/CVF In- ternational Conference on Computer Vision (ICCV) , pages 32...

    2021

  41. [48]

    Geoudf: Surface reconstruction from 3D point clouds via geometry-guided distance representation

    Siyu Ren, Junhui Hou, Xiaodong Chen, Ying He, and Wen- ping Wang. Geoudf: Surface reconstruction from 3D point clouds via geometry-guided distance representation. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pages 14214–14224, 2023. 3

    2023

  42. [49]

    PointR- CNN: 3D object proposal generation and detection from 10 point cloud

    Shaoshuai Shi, Xiaogang Wang, and Hongsheng Li. PointR- CNN: 3D object proposal generation and detection from 10 point cloud. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–779, 2019. 3

    2019

  43. [50]

    Very deep convo- lutional networks for large-scale image recognition

    Karen Simonyan and Andrew Zisserman. Very deep convo- lutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556, 2014. 5

    2014 arXiv

  44. [51]

    Sun rgb-d: A rgb-d scene understanding benchmark suite

    Shuran Song, Samuel P Lichtenberg, and Jianxiong Xiao. Sun rgb-d: A rgb-d scene understanding benchmark suite. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 567–576, 2015. 5

    2015

  45. [52]

    Kpconv: Flexible and deformable convolution for point clouds

    Hugues Thomas, Charles R Qi, Jean-Emmanuel Deschaud, Beatriz Marcotegui, Franc ¸ois Goulette, and Leonidas J Guibas. Kpconv: Flexible and deformable convolution for point clouds. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pages 6411–6420,

  46. [53]

    Geo- mae: Masked geometric target prediction for self-supervised point cloud pre-training

    Xiaoyu Tian, Haoxi Ran, Yue Wang, and Hang Zhao. Geo- mae: Masked geometric target prediction for self-supervised point cloud pre-training. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 13570–13580, 2023. 2

    2023

  47. [54]

    Neus: Learning neural implicit surfaces by volume rendering for multi-view reconstruction

    Peng Wang, Lingjie Liu, Yuan Liu, Christian Theobalt, Taku Komura, and Wenping Wang. Neus: Learning neural implicit surfaces by volume rendering for multi-view reconstruction. arXiv preprint arXiv:2106.10689, 2021. 3

    2021 arXiv

  48. [55]

    Freesplat: Generalizable 3D gaussian splatting to- wards free-view synthesis of indoor scenes

    Yunsong Wang, Tianxin Huang, Hanlin Chen, and Gim Hee Lee. Freesplat: Generalizable 3D gaussian splatting to- wards free-view synthesis of indoor scenes. arXiv preprint arXiv:2405.17958, 2024. 3, 4

    2024 arXiv

  49. [56]

    Masked scene contrast: A scalable framework for unsu- pervised 3D representation learning

    Xiaoyang Wu, Xin Wen, Xihui Liu, and Hengshuang Zhao. Masked scene contrast: A scalable framework for unsu- pervised 3D representation learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 9415–9424, 2023. 2

    2023

  50. [58]

    Point cloud pre- training with natural 3D structures

    Ryosuke Yamada, Hirokatsu Kataoka, Naoya Chiba, Yukiyasu Domae, and Tetsuya Ogata. Point cloud pre- training with natural 3D structures. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 21283–21293, 2022. 6

    2022

  51. [59]

    Im- plicit autoencoder for point-cloud self-supervised represen- tation learning

    Siming Yan, Zhenpei Yang, Haoxiang Li, Chen Song, Li Guan, Hao Kang, Gang Hua, and Qixing Huang. Im- plicit autoencoder for point-cloud self-supervised represen- tation learning. In Proceedings of the IEEE/CVF Inter- national Conference on Computer Vision (ICCV) , pages 14530–...

    2023

  52. [60]

    Gd-mae: gen- erative decoder for mae pre-training on lidar point clouds

    Honghui Yang, Tong He, Jiaheng Liu, Hua Chen, Boxi Wu, Binbin Lin, Xiaofei He, and Wanli Ouyang. Gd-mae: gen- erative decoder for mae pre-training on lidar point clouds. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pages 9403–9414,

  53. [61]

    Pred: pre-training via semantic rendering on lidar point clouds

    Hao Yang, Haiyang Wang, Di Dai, and Liwei Wang. Pred: pre-training via semantic rendering on lidar point clouds. Proceedings of the Advances in Neural Information Process- ing Systems (NeurIPS), 36, 2024. 3

    2024

  54. [62]

    Unipad: A universal pre-training paradigm for autonomous driving

    Honghui Yang, Sha Zhang, Di Huang, Xiaoyang Wu, Haoyi Zhu, Tong He, Shixiang Tang, Hengshuang Zhao, Qibo Qiu, Binbin Lin, et al. Unipad: A universal pre-training paradigm for autonomous driving. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogniti...

    2024

  55. [63]

    3DSSD: Point-based 3D single stage object detector

    Zetong Yang, Yanan Sun, Shu Liu, and Jiaya Jia. 3DSSD: Point-based 3D single stage object detector. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition (CVPR), pages 11040–11048, 2020. 3

    2020

  56. [64]

    V ol- ume rendering of neural implicit surfaces

    Lior Yariv, Jiatao Gu, Yoni Kasten, and Yaron Lipman. V ol- ume rendering of neural implicit surfaces. In Proceedings of the Advances in Neural Information Processing Systems (NeurIPS), pages 4805–4815, 2021. 3

    2021

  57. [65]

    Gspn: Generative shape proposal network for 3D instance segmentation in point cloud

    Li Yi, Wang Zhao, He Wang, Minhyuk Sung, and Leonidas J Guibas. Gspn: Generative shape proposal network for 3D instance segmentation in point cloud. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 3947–3956, 2019. 7

    2019

  58. [66]

    Point-bert: Pre-training 3D point cloud transformers with masked point modeling

    Xumin Yu, Lulu Tang, Yongming Rao, Tiejun Huang, Jie Zhou, and Jiwen Lu. Point-bert: Pre-training 3D point cloud transformers with masked point modeling. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 19313–19322, 2022. 6

    2022

  59. [67]

    Mip-splatting: Alias-free 3D gaussian splat- ting

    Zehao Yu, Anpei Chen, Binbin Huang, Torsten Sattler, and Andreas Geiger. Mip-splatting: Alias-free 3D gaussian splat- ting. In Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR), pages 19447– 19456, 2024. 3

    2024

  60. [68]

    The unreasonable effectiveness of deep features as a perceptual metric

    Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shecht- man, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion (CVPR), pages 586–595, 2018. 5

    2018

  61. [69]

    Point-m2ae: multi-scale masked autoencoders for hierarchical point cloud pre-training

    Renrui Zhang, Ziyu Guo, Peng Gao, Rongyao Fang, Bin Zhao, Dong Wang, Yu Qiao, and Hongsheng Li. Point-m2ae: multi-scale masked autoencoders for hierarchical point cloud pre-training. Proceedings of the Advances in Neural In- formation Processing Systems (NeurIPS) , 35:27061–27074,

  62. [70]

    H3dnet: 3D object detection using hybrid geometric prim- itives

    Zaiwei Zhang, Bo Sun, Haitao Yang, and Qixing Huang. H3dnet: 3D object detection using hybrid geometric prim- itives. In Proceedings of the European Conference on Com- puter Vision (ECCV), pages 311–329. Springer, 2020. 6

    2020

  63. [71]

    Self-supervised pretraining of 3D features on any point-cloud

    Zaiwei Zhang, Rohit Girdhar, Armand Joulin, and Ishan Misra. Self-supervised pretraining of 3D features on any point-cloud. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) , pages 10252– 10263, 2021. 2, 6

    2021

  64. [72]

    Point transformer

    Hengshuang Zhao, Li Jiang, Jiaya Jia, Philip HS Torr, and Vladlen Koltun. Point transformer. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pages 16259–16268, 2021. 7, 13

    2021

  65. [73]

    V oxelNet: End-to-end learning for point cloud based 3D object detection

    Yin Zhou and Oncel Tuzel. V oxelNet: End-to-end learning for point cloud based 3D object detection. In Proceedings of 11 the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 4490–4499, 2018. 3

    2018

  66. [74]

    Ponderv2: Pave the way for 3D foundataion model with a universal pre-training paradigm

    Haoyi Zhu, Honghui Yang, Xiaoyang Wu, Di Huang, Sha Zhang, Xianglong He, Tong He, Hengshuang Zhao, Chun- hua Shen, Yu Qiao, et al. Ponderv2: Pave the way for 3D foundataion model with a universal pre-training paradigm. arXiv preprint arXiv:2310.08586, 2023. 2 A.1. Visualizatio...

    arXiv

  67. [621]

    Springer, 2022. 1, 2

    2022

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.