Pith. sign in

REVIEW 4 major objections 6 minor 35 references

Face recognition on point cloud with cgan-top for denoising

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read One denoiser lifts noisy 3D face recognition by 14.81%.

desk verdict The three-plane cGAN denoising idea is sensible, but the headline 14.81% gain is not supported because no LDGCNN-without-denoising baseline is reported. read the letter →

arxiv 2506.06864 v1 pith:U277ETLN submitted 2025-06-07 cs.CV cs.AI

Junyu Liu , Jianfeng Ren , Sunhong Liang , Xudong Jiang This is my paper
classification cs.CVcs.AI
keywords 3DfacerecognitionpointclouddenoisingconditionalGANthreeorthogonalplanesLDGCNNBosphorusdatasetnoisyend-to-endlearning
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 proposes an end-to-end system that first denoises a noisy 3D face point cloud and then recognizes the identity from the cleaned cloud. The denoiser, cGAN-TOP, projects the cloud onto three orthogonal planes, treats the coordinate perpendicular to each plane as a gray value, and runs a conditional GAN per plane to remove noise. A Recognition Feature Discriminator pushes the denoised cloud to retain identity-discriminative features, while a Visual Appearance Discriminator keeps it geometrically clean. On the Bosphorus dataset the method beats the previous baselines at every tested noise level, with the largest gain of 14.81% accuracy at the highest noise level under the Random Setting.

What carries the argument

cGAN-TOP (Conditional GAN on Three Orthogonal Planes): the point cloud is projected onto xy, yz, and zx planes; in each plane the orthogonal coordinate becomes the gray value and a UNet-style generator denoises that 2D image. Two discriminators guide training: a Visual Appearance Discriminator (DCGAN-like) that enforces geometric cleanness and a Recognition Feature Discriminator that enforces identity-relevant features. The denoised coordinates are mapped back and combined into a point cloud, which is then classified by an adapted LDGCNN that links hierarchical local and multi-scale neighboring features.

What would settle it

Train the same LDGCNN recognizer on the noisy point clouds without any denoising, using the identical training protocol; if its accuracy matches or exceeds the cGAN-TOP pipeline at σ2=64 in the Random Setting, the denoiser is not the cause of the 14.81% gain, and also inspect the reconstruction step to see whether multiple points projected into the same pixel are handled without collapse or duplication.

Watch

Extended reading notes

Core claim

The central claim is that denoising and recognition should be trained together rather than treated as separate stages, and that denoising can be done effectively on three orthogonal 2D projections of a point cloud by treating each spatial coordinate as a pixel gray value. In this design, cGAN-TOP removes noise in each coordinate channel, and the two discriminators together guarantee both visual cleanness and recognition suitability of the output. An adapted LDGCNN then classifies the reconstructed point cloud. The paper reports that this integrated pipeline outperforms LeD3D and FER under all five Gaussian noise levels on Bosphorus, with accuracy gains up to 14.81% at σ2=64 in the Random Setting, and also produces lower Chamfer and Point-to-Mesh distances than PCNet, DMR, and SB at high noise levels.

Load-bearing premise

The reported improvement assumes that mapping the denoised 2D planes back into a sparse 3D point cloud via the unspecified inverse function preserves enough geometric structure for the recognizer, and that the measured gain comes from the cGAN-TOP denoiser rather than from the stronger LDGCNN backbone alone.

Editorial extensions

If this is right

  • If the cGAN-TOP denoiser is the source of the gain, then joint denoising-recognition training on orthogonal projections is a viable general recipe for other sparse 3D recognition tasks on noisy sensor data.
  • The Recognition Feature Discriminator is what ties denoising to recognition; the ablation suggests dropping either discriminator costs accuracy, so recognition-aware denoising is necessary for the reported results.
  • At high noise levels the method improves both recognition accuracy and geometric cleanness, indicating the two objectives are not in conflict in this framework.
  • The paper claims the approach suits thin, surface-like point clouds where 3D graph convolutions struggle, such as facial scans from consumer depth sensors.

Reading between the lines

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

  • The paper does not report recognition with LDGCNN alone on noisy clouds; an obvious test is whether the 14.81% gain survives when the denoiser is removed and the same backbone is used, which would isolate the denoiser's contribution.
  • The inverse mapping from denoised 2D images back to a 3D point cloud is not specified for pixels receiving multiple points or no points; resolving this ambiguity could affect the geometry-preservation claim.
  • If the same cGAN-TOP design works for other sparse 3D recognition problems, such as object class recognition on LiDAR-style clouds, the orthogonal-projection denoising idea becomes a general preprocessing layer.
  • The reported gains are on one dataset with injected Gaussian noise; testing on real sensor noise from depth-camera scans would show whether the advantage transfers.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper proposes an end-to-end 3D face recognition pipeline for noisy point clouds, consisting of a conditional GAN that denoises three orthogonal plane projections of the point cloud (cGAN-TOP) followed by an LDGCNN classifier. The method is evaluated on the Bosphorus dataset under five Gaussian noise levels and two train/test settings. The authors report accuracy improvements of up to 14.81% over prior methods and better Chamfer/Point-to-Mesh distances against dedicated denoising models.

Significance. If the claims are substantiated, the work would offer a practical way to leverage 2D GANs for point-cloud denoising in a recognition-driven loop, with a plausible dual-discriminator design (visual appearance plus recognition features). The paper explicitly targets an end-to-end integration, which is a worthwhile direction. However, the current evidence does not establish that the proposed denoiser is responsible for the reported gains, and the inverse mapping from denoised images back to point clouds is left unspecified. The manuscript also lacks reproducibility details such as error bars, training hyper-parameters, and precise noise injection protocols. These issues are load-bearing for the central claim.

major comments (4)
  1. [Section 3.2, Table 1] The central quantitative claim of a maximum 14.81% accuracy gain at σ2=64 under the Random Setting is supported only by comparisons with LeD3D [16] and FER [15]. Since LDGCNN is a strong point-cloud recognition backbone, the reported advantage may be caused by the backbone rather than by the proposed cGAN-TOP denoiser. No control experiment is reported in which the same LDGCNN is trained and evaluated on noisy point clouds without the denoising stage. Please add this condition and report mean ± standard deviation over multiple runs with appropriate significance tests.
  2. [Section 2.2, Fig. 2] The reconstruction step, in which the three denoised 2D images are mapped back to a denoised 3D point cloud via the inverse function f′, is not specified. In particular, the paper does not explain how points that project to the same pixel in a given plane are resolved, how the three independently denoised coordinate images are fused into a consistent point set, or what happens when the denoised images produce different point counts in the three planes. This missing definition makes the method irreproducible and raises the possibility of systematic loss of geometric information. Please provide a precise algorithm for f′ and discuss its behavior with respect to occlusions, multiplicity, and information preservation.
  3. [Section 2.2 and Table 3] The ablation study varies which discriminator losses are used, but it never removes the cGAN-TOP denoiser entirely; feeding noisy point clouds directly to LDGCNN under identical training conditions is not reported. Consequently, the experiment does not isolate the contribution of the denoising module. In addition, the hyper-parameters λ1=0.67 and λ2=0.33 are said to be tuned to achieve optimal performance on the same dataset, but no validation split or sensitivity analysis is described. Please clarify how these values were selected and report accuracy over a range of λ1/λ2 values to rule out overfitting to the test set.
  4. [Section 3.1] The experimental protocol reports only single-run accuracy numbers without standard deviations or the number of repeated trials. For several noise levels the differences between methods are small (e.g., Neutral Setting at σ2=4: Ours 98.74 vs FER 96.88), so it is impossible to assess whether the observed improvements are statistically reliable. Please provide mean ± std over at least three independent runs, or explicitly state whether training is deterministic and the test set is fixed so that repeated runs are unnecessary. The paper also omits details on the number of training/test subjects, the exact noise injection procedure (e.g., per-coordinate independent Gaussian noise or along surface normals), and the point-cloud alignment protocol.
minor comments (6)
  1. [Abstract] The abstract reports a maximum gain of 14.81% without stating that it occurs at σ2=64 under the Random Setting; please make this condition explicit.
  2. [Section 2.2] The notation f: (x,y) → (x′, y′) is introduced but never used or explained; please clarify what coordinate transformation this represents and how it relates to the pixel grid.
  3. [Section 2.1] The phrase 'the three denoised coordinates are then combined' is ambiguous in light of the missing inverse mapping; consider rewriting once the reconstruction procedure is specified.
  4. [Table 2] Please specify how Chamfer distance and Point-to-Mesh distance are computed, including point-density normalization and whether the denoised point clouds are subsampled or resized to a common resolution.
  5. [Section 3.2] The text says the proposed method 'shows advantages at a high noise level and comparable performance to SB at a low noise level,' but Table 2 does not include confidence intervals or significance tests for the CD/P2M metrics; please address this.
  6. [Throughout] There are minor typographical and consistency issues: 'VFD' and 'LRFD' in Table 3 do not match the 'VAD' and 'RFD' abbreviations used elsewhere; 'visually' is misspelled as 'visual' in Section 2.2; and 'plan' appears instead of 'plane' in the conclusion.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the reported gain is an empirical comparison, not a derivation, and the self-citations are background only.

full rationale

The paper does not contain a derivation chain that could reduce to its inputs. cGAN-TOP denoises the point cloud by treating each orthogonal coordinate as a gray value and reconstructing the point cloud, after which an adapted LDGCNN performs recognition; the two modules are distinct, and the claimed 14.81% improvement is an empirical accuracy comparison on the Bosphorus dataset against LeD3D and FER (Table 1). No equation or definition in the paper makes the reported accuracy equal to an input by construction. The hyperparameters lambda1=0.67 and lambda2=0.33 are said to be tuned for optimal performance (Section 2.2), which is a potential overfitting or leakage concern if the tuning set overlaps the test set, but the accuracy is not defined as the loss weight, so this is not a circular reduction. Self-citations (e.g., Ref. [20] on noise-resistant LBP and Ref. [3] on 3D pose estimation) appear only in background references and are not load-bearing. The missing LDGCNN-without-denoising control and the unspecified inverse mapping f' are experimental-design and reproducibility gaps, not circularity; they affect attribution of the gain and reproducibility but do not make any prediction equivalent to its inputs. Overall, no significant circularity.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

No new physical or conceptual entities are introduced; the cGAN-TOP is an architectural arrangement of existing GAN and CNN components.

free parameters (2)
  • λ1 = 0.67
    Weight for recognition feature discriminator loss, tuned to achieve optimal performance on the Bosphorus dataset (Section 2.2).
  • λ2 = 0.33
    Weight for visual appearance discriminator loss, tuned jointly with λ1 on the same dataset (Section 2.2).
assumptions (4)
  • domain assumption Bosphorus dataset with Gaussian noise σ² ∈ {4,8,16,32,64} represents realistic noisy 3D face recognition conditions.
    Used to define the experimental protocol in Section 3.1.
  • ad hoc to paper Projecting a point cloud onto three orthogonal planes and treating the orthogonal coordinate as a gray value preserves enough information for denoising and recognition.
    This is the core design premise of cGAN-TOP, stated in Section 2.2 without proof or analysis of occlusion or collisions.
  • ad hoc to paper The inverse mapping f' from structured denoised images back to sparse point clouds is well-defined and lossless.
    Invoked in Section 2.2 but never specified; multiple points may project to one pixel.
  • domain assumption LDGCNN, originally designed for general point cloud classification, transfers to face recognition with only channel-count modifications.
    Adaptation described in Section 2.3.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Face recognition on point cloud with cgan-top for denoising." pith.science (2026). https://pith.science/paper/U277ETLN

@misc{pith2026250606864,
  author       = {Pith},
  title        = {Pith review of: Face recognition on point cloud with cgan-top for denoising},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/U277ETLN}},
  note         = {Machine review of arXiv:2506.06864}
}
read the original abstract

Face recognition using 3D point clouds is gaining growing interest, while raw point clouds often contain a significant amount of noise due to imperfect sensors. In this paper, an end-to-end 3D face recognition on a noisy point cloud is proposed, which synergistically integrates the denoising and recognition modules. Specifically, a Conditional Generative Adversarial Network on Three Orthogonal Planes (cGAN-TOP) is designed to effectively remove the noise in the point cloud, and recover the underlying features for subsequent recognition. A Linked Dynamic Graph Convolutional Neural Network (LDGCNN) is then adapted to recognize faces from the processed point cloud, which hierarchically links both the local point features and neighboring features of multiple scales. The proposed method is validated on the Bosphorus dataset. It significantly improves the recognition accuracy under all noise settings, with a maximum gain of 14.81%.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

35 extracted references · 32 canonical work pages

  1. [16]

    Dynamic graph CNN for learning on point clouds,

    Y . Wang, Y . Sun, Z. Liu, S. E. Sarma, M. M. Bronstein, and J. M. Solomon, “Dynamic graph CNN for learning on point clouds,”CoRR, vol. abs/1801.07829, 2018

    arXiv 2018

  2. [15]

    VERAM: view-enhanced recurrent attention model for 3D shape classifi- cation,

    S. Chen, L. Zheng, Y . Zhang, Z. Sun, and K. Xu, “VERAM: view-enhanced recurrent attention model for 3D shape classifi- cation,”IEEE Trans. Vis. Comput. Graph., vol. 25, no. 12, pp. 3244–3257, 2019

    work page 2019

  3. [1]

    Face recognition on point cloud with cgan-top for denoising

    INTRODUCTION 3D face recognition becomes increasing popular, as 3D data provide more information and are less affected by illumina- tion [1–5]. Among all 3D models, point clouds are most com- monly used [6]. Early methods for 3D face recognition often manually specify geometric properties for feature extraction [7]. 2D deep learning methods [8–10] project...

    work page Pith review arXiv 2025

  4. [2]

    Overview of Proposed Method Point clouds are sparse and unstructured [12] (unlike densely structured models such as 2D images), enabling them to rep- resent complex models flexibly

    PROPOSED METHOD 2.1. Overview of Proposed Method Point clouds are sparse and unstructured [12] (unlike densely structured models such as 2D images), enabling them to rep- resent complex models flexibly. But recognition using 3D point clouds faces two main challenges. 1) The sparsity poses a challenge to point cloud processing since traditional 2D and 3D C...

    work page

  5. [3]

    3D Graph Convolutional Networks can’t effectively pro- cess noisy 3D face models due to the thin surface-like char- acteristics of face point clouds. Therefore, we apply three orthogonal projections to decouple(x, y, z)values of a point cloud into three orthogonal planes and take the coordinate or- thogonal to the plane as the gray value of the point in t...

    work page

  6. [4]

    Experimental Settings The proposed method is evaluated on the Bosphorus dataset [19], a commonly used 3D face point cloud dataset

    EXPERIMENTAL RESULTS 3.1. Experimental Settings The proposed method is evaluated on the Bosphorus dataset [19], a commonly used 3D face point cloud dataset. It con- tains 45,000 3D faces of 150 individuals, including variations in expression, occlusion, and pose. We use two common settings [15]. 1)Neutral Setting: Faces with a neutral ex- pression (60%of ...

    work page

  7. [5]

    CONCLUSION In this paper, we propose an end-to-end 3D face point cloud denoising and recognition model to provide both high recog- nition accuracy and visually clean point clouds. The proposed cGAN-TOP provides an efficient and effective way to directly remove the noise in the 3D coordinates of the point cloud by using cGAN on each of the three orthogonal...

    work page

  8. [6]

    Learning from millions of 3D scans for large-scale 3D face recognition,

    S. Gilani and A. Mian, “Learning from millions of 3D scans for large-scale 3D face recognition,” inIEEE Conference on Com- puter Vision and Pattern Recognition, 2018, pp. 1896–1905

    work page 2018

Show all 35 references
  1. [7]

    PointFace: Point set based feature learning for 3D face recognition,

    C. Jiang, S. Lin, W. Chen, F. Liu, and L. Shen, “PointFace: Point set based feature learning for 3D face recognition,” in International IEEE Joint Conference on Biometrics, 2021, pp. 1–8

    2021

  2. [8]

    Feature boosting network for 3D pose estima- tion,

    J. Liu, H. Ding, A. Shahroudy, L. Duan, X. Jiang, G. Wang, and A. Kot, “Feature boosting network for 3D pose estima- tion,”IEEE transactions on pattern analysis and machine in- telligence, vol. 42, no. 2, pp. 494–501, 2019

    2019

  3. [9]

    Blood vessel segmentation from fundus image by a cascade classification framework,

    X. Wang, X. Jiang, and J. Ren, “Blood vessel segmentation from fundus image by a cascade classification framework,” Pattern Recognition, vol. 88, pp. 331–341, 2019

    2019

  4. [10]

    Face2statistics: user-friendly, low-cost and effective alternative to in-vehicle sensors/monitors for drivers,

    Z. Xiong, J. Wang, W. Jin, J. Liu, Y . Duan, Z. Song, and X. Peng, “Face2statistics: user-friendly, low-cost and effective alternative to in-vehicle sensors/monitors for drivers,” inInter- national Conference on Human-Computer Interaction, 2022

    2022

  5. [11]

    Linked dynamic graph CNN: learning on point cloud via linking hier- archical features,

    K. Zhang, M. Hao, J. Wang, C. W. de Silva, and C. Fu, “Linked dynamic graph CNN: learning on point cloud via linking hier- archical features,”CoRR, vol. abs/1904.10014, 2019

    1904 arXiv

  6. [12]

    Shape matching and object recognition using shape contexts,

    S. J. Belongie, J. Malik, and J. Puzicha, “Shape matching and object recognition using shape contexts,”IEEE Trans. Pattern Anal. Mach. Intell., vol. 24, no. 4, pp. 509–522, 2002

    2002

  7. [13]

    3D ShapeNets: A deep representation for volumetric shapes,

    Z. Wu, S. Song, A. Khosla, F. Yu, L. Zhang, X. Tang, and J. Xiao, “3D ShapeNets: A deep representation for volumetric shapes,” inIEEE Conference on Computer Vision and Pattern Recognition, 2015, pp. 1912–1920

    2015

  8. [14]

    SPNet: Deep 3D ob- ject classification and retrieval using stereographic projection,

    M. Yavartanoo, E. Kim, and K. M. Lee, “SPNet: Deep 3D ob- ject classification and retrieval using stereographic projection,” inAsian Conference on Computer Vision, 2018, pp. 691–706

    2018

  9. [17]

    PointNet: A 3D convolutional neural network for real-time object class recog- nition,

    A. Garcia-Garcia, F. Gomez-Donoso, J. G. Rodr´ıguez, S. Orts- Escolano, M. Cazorla, and J. Azor´ın L´opez, “PointNet: A 3D convolutional neural network for real-time object class recog- nition,” inInternational Joint Conference on Neural Networks, 2016, pp. 1578–1584

    2016

  10. [18]

    Learning directly from synthetic point clouds for

    Z. Zhang, F. Da, and Y . Yu, “Learning directly from synthetic point clouds for ”in-the-wild” 3D face recognition,”Pattern Recognit., vol. 123, pp. 108394, 2022

    2022

  11. [19]

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

    C. R. Qi, L. Yi, H. Su, and L. J. Guibas, “Pointnet++: Deep hierarchical feature learning on point sets in a metric space,” inAdvances in Neural Information Processing Systems, 2017, pp. 5099–5108

    2017

  12. [20]

    Learning flexibly dis- tributional representation for low-quality 3D face recognition,

    Z. Zhang, C. Yu, S. Xu, and H. Li, “Learning flexibly dis- tributional representation for low-quality 3D face recognition,” inAAAI Conference on Artificial Intelligence, 2021, pp. 3465– 3473

    2021

  13. [21]

    Led3D: A lightweight and efficient deep approach to recognizing low- quality 3D faces,

    G. Mu, D. Huang, G. Hu, J. Sun, and Y . Wang, “Led3D: A lightweight and efficient deep approach to recognizing low- quality 3D faces,” inIEEE Conference on Computer Vision and Pattern Recognition, 2019, pp. 5773–5782

    2019

  14. [22]

    Patch-based progressive 3D point set upsampling,

    Y . Wang, S. Wu, H. Huang, D. Cohen-Or, and O. Sorkine- Hornung, “Patch-based progressive 3D point set upsampling,” inIEEE Conference on Computer Vision and Pattern Recogni- tion, 2019, pp. 5958–5967

    2019

  15. [23]

    Rain- component-aware capsule-GAN for single image de-raining,

    F. Yang, J. Ren, Z. Lu, J. Zhang, and Q. Zhang, “Rain- component-aware capsule-GAN for single image de-raining,” Pattern Recognition, vol. 123, pp. 108377, 2022

    2022

  16. [24]

    Bosphorus database for 3D face analysis,

    A. Savran, N. Aly ¨uz, H. Dibeklioglu, O. C ¸ eliktutan, B. G ¨okberk, B. Sankur, and L. Akarun, “Bosphorus database for 3D face analysis,” inBiometrics and Identity Management, 2008, vol. 5372, pp. 47–56

    2008

  17. [25]

    Noise-resistant local binary pattern with an embedded error-correction mechanism,

    J. Ren, X. Jiang, and J. Yuan, “Noise-resistant local binary pattern with an embedded error-correction mechanism,”IEEE Transactions on Image Processing, vol. 22, no. 10, pp. 4049– 4060, 2013

    2013

  18. [26]

    Repcd- net: Feature-aware recurrent point cloud denoising network,

    H. Chen, Z. Wei, X. Li, Y . Xu, M. Wei, and J. Wang, “Repcd- net: Feature-aware recurrent point cloud denoising network,” International Journal of Computer Vision, vol. 130, no. 3, pp. 615–629, 2022

    2022

  19. [27]

    Three-dimensional point cloud denoising via a gravitational feature function,

    C. Shi, C. Wang, X. Liu, S. Sun, B. Xiao, X. Li, and G. Li, “Three-dimensional point cloud denoising via a gravitational feature function,”Applied Optics, vol. 61, no. 6, pp. 1331– 1343, 2022

    2022

  20. [28]

    Spatial- context-aware deep neural network for multi-class image clas- sification,

    J. Zhang, Q. Zhang, J. Ren, Y . Zhao, and J. Liu, “Spatial- context-aware deep neural network for multi-class image clas- sification,” inIEEE International Conference on Acoustics, Speech and Signal Processing, 2022, pp. 1960–1964

    2022

  21. [29]

    Attention-based dual- stream vision transformer for radar gait recognition,

    S. Chen, W. He, J. Ren, and X. Jiang, “Attention-based dual- stream vision transformer for radar gait recognition,” inIEEE International Conference on Acoustics, Speech and Signal Pro- cessing, 2022, pp. 3668–3672

    2022

  22. [30]

    Hierarchi- cal convit with attention-based relational reasoner for visual analogical reasoning,

    W. He, J. Zhang, J. Ren, R. Bai, and X. Jiang, “Hierarchi- cal convit with attention-based relational reasoner for visual analogical reasoning,” inAAAI Conference on Artificial Intel- ligence, 2023

    2023

  23. [31]

    U-net: Convolu- tional networks for biomedical image segmentation,

    O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolu- tional networks for biomedical image segmentation,” inMed- ical Image Computing and Computer-Assisted Intervention, 2015, vol. 9351, pp. 234–241

    2015

  24. [32]

    Unsupervised represen- tation learning with deep convolutional generative adversarial networks,

    A. Radford, L. Metz, and S. Chintala, “Unsupervised represen- tation learning with deep convolutional generative adversarial networks,”arXiv preprint arXiv:1511.06434, 2015

    2015 arXiv

  25. [33]

    Pointcleannet: Learning to denoise and re- move outliers from dense point clouds,

    M. Rakotosaona, V . La Barbera, P. Guerrero, N. J. Mitra, and M. Ovsjanikov, “Pointcleannet: Learning to denoise and re- move outliers from dense point clouds,” inComputer Graphics Forum, 2020, vol. 39, pp. 185–203

    2020

  26. [34]

    Differentiable manifold reconstruction for point cloud denoising,

    S. Luo and W. Hu, “Differentiable manifold reconstruction for point cloud denoising,” inACM international conference on multimedia, 2020, pp. 1330–1338

    2020

  27. [35]

    Score-based point cloud denoising,

    S. Luo and W. Hu, “Score-based point cloud denoising,” in IEEE International Conference on Computer Vision, 2021, pp. 4583–4592

    2021

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.