pith. sign in

arxiv: 2404.07106 · v2 · pith:42JABDQUnew · submitted 2024-04-10 · 💻 cs.CV · cs.GR

3DMambaComplete: Exploring Structured State Space Model for Point Cloud Completion

Yixuan Li , Weidong Yang , Ben Fei This is my paper

Pith reviewed 2026-05-24 02:06 UTC · model grok-4.3

classification 💻 cs.CV cs.GR
keywords point cloud completionMambastructured state space modelHyperPoint3D reconstructionTransformer replacementpoint cloud generation
0
0 comments X

The pith

Mamba's selection mechanism completes point clouds by encoding features without pooling-induced detail loss.

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

The paper aims to show that a structured state space model can replace Transformer attention and pooling for point cloud completion, yielding higher-fidelity outputs from incomplete inputs. It builds a pipeline around Mamba to generate, spread, and deform HyperPoints into full 3D structures. A reader would care because many real-world scans from sensors are partial, and current methods either drop local geometry or scale poorly with sequence length. The approach targets both accuracy and efficiency on long point sets. Experiments across benchmarks report better quantitative metrics and visual results than prior state-of-the-art networks.

Core claim

3DMambaComplete encodes incomplete point clouds with Mamba's selection mechanism inside the HyperPoint Generation module to predict a compact set of HyperPoints, then applies a spread operation to disperse them spatially and a deformation step that converts their 2D mesh representation into a dense 3D point cloud; this pipeline is claimed to avoid the local-detail erosion typical of pooling while keeping computation lower than attention, resulting in outputs that exceed existing methods on standard completion benchmarks.

What carries the argument

Mamba selection mechanism inside HyperPoint Generation, which encodes features and predicts HyperPoints that are then spread and deformed into the final 3D structure.

If this is right

  • Point cloud sequences can be processed at lower computational cost while retaining local detail.
  • Downsampled HyperPoints serve as an effective intermediate representation for 3D reconstruction.
  • The deformation of 2D meshes into 3D points produces denser and more accurate completions than direct upsampling.
  • Qualitative and quantitative gains hold across multiple established point-cloud benchmarks.

Where Pith is reading between the lines

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

  • The same Mamba-plus-HyperPoint pattern may transfer to other 3D tasks such as denoising or upsampling.
  • If the efficiency gains scale, the method could support completion on edge devices with limited memory.
  • State-space selection offers a general alternative to attention when global context must be modeled without explicit pairwise comparisons.

Load-bearing premise

Mamba's selection mechanism together with the HyperPoint spread and deformation steps can capture and reconstruct fine local geometry without the losses introduced by pooling in Transformer pipelines.

What would settle it

On the PCN or Completion3D benchmark, if the Chamfer distance or F-Score of 3DMambaComplete does not exceed that of the strongest published Transformer completion model under identical training settings.

Figures

Figures reproduced from arXiv: 2404.07106 by Ben Fei, Weidong Yang, Yixuan Li.

Figure 1
Figure 1. Figure 1: An overview of 3DMambaComplete. 3DMambaComplete mainly consists of three modules: (1) Given an incomplete [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Details of HyperPoint Generation. The output can be calculated by Z ′ 𝑝 = DW (MLP (LN (Zp))) (4) Z ′′ 𝑝 = MLP (LN (SSM(𝜎(Z′ p )))) (5) Z𝑝 = Z ′′ 𝑝 × 𝜎(𝐿𝑁 (Z𝑝−1)) + Z𝑝−1 (6) where Z𝑝 ∈ 𝑅 𝑁 × (𝐶+3) taken from the sampled point coordinates and features. The SSM is the key module in the mamba block, with a detailed description provided in Section 3.2.1. The enhanced sam￾pled point feature values are later obta… view at source ↗
Figure 3
Figure 3. Figure 3: Detailed structure for HyperPoint Spread Module. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: The pipeline of the Point Deformation [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Visualization comparison of point cloud completion on PCN [51] dataset. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: Visualization comparison of point cloud completion on ShapeNet55 [49] dataset. [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Visualization comparison of point cloud completion on ShapeNetUnseen21 [49] dataset. [PITH_FULL_IMAGE:figures/full_fig_p008_8.png] view at source ↗
read the original abstract

Point cloud completion aims to generate a complete and high-fidelity point cloud from an initially incomplete and low-quality input. A prevalent strategy involves leveraging Transformer-based models to encode global features and facilitate the reconstruction process. However, the adoption of pooling operations to obtain global feature representations often results in the loss of local details within the point cloud. Moreover, the attention mechanism inherent in Transformers introduces additional computational complexity, rendering it challenging to handle long sequences effectively. To address these issues, we propose 3DMambaComplete, a point cloud completion network built on the novel Mamba framework. It comprises three modules: HyperPoint Generation encodes point cloud features using Mamba's selection mechanism and predicts a set of Hyperpoints. A specific offset is estimated, and the down-sampled points become HyperPoints. The HyperPoint Spread module disperses these HyperPoints across different spatial locations to avoid concentration. Finally, a deformation method transforms the 2D mesh representation of HyperPoints into a fine-grained 3D structure for point cloud reconstruction. Extensive experiments conducted on various established benchmarks demonstrate that 3DMambaComplete surpasses state-of-the-art point cloud completion methods, as confirmed by qualitative and quantitative analyses.

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 paper introduces 3DMambaComplete, a point cloud completion architecture that replaces Transformer-based global feature encoding with the Mamba structured state space model. The network consists of a HyperPoint Generation module that uses Mamba's input-dependent selection to encode features, estimate offsets, and downsample to a set of HyperPoints; a HyperPoint Spread module that disperses these points spatially; and a final deformation stage that converts a 2D mesh representation of the HyperPoints into a dense 3D point cloud. The central claim is that this design avoids the local-detail loss associated with pooling in Transformers while achieving linear complexity, and that extensive experiments on established benchmarks show quantitative and qualitative superiority over prior state-of-the-art methods.

Significance. If the experimental superiority and the claimed preservation of local geometry both hold, the work would constitute the first demonstration of Mamba-based point cloud completion and would supply a concrete efficiency argument (linear vs. quadratic complexity) together with a new set of architectural primitives (HyperPoint generation, spread, and 2D-to-3D deformation). These elements could influence subsequent work on long-sequence 3D tasks. The significance is currently limited by the absence of any reported dataset names, metrics, baseline implementations, ablation studies, or error analysis that would allow independent verification of the performance claims.

major comments (1)
  1. [Abstract / HyperPoint Generation module] Abstract (and the description of the HyperPoint Generation module): the central motivation states that pooling operations in Transformers cause loss of local details, yet the proposed pipeline explicitly downsamples the input after offset estimation to produce the HyperPoints on which Mamba operates. No analysis is supplied showing that this downsampling step preserves the fine-grained geometry that the method claims to protect; if the downsampling discards information before or during Mamba encoding, the claimed advantage over pooling-based Transformers does not follow.
minor comments (2)
  1. [Abstract] The abstract asserts superiority on 'various established benchmarks' but supplies no concrete dataset names, metrics (e.g., CD, EMD, F-score), baseline list, or quantitative tables; these details are required for any performance claim.
  2. The terms 'HyperPoints', 'HyperPoint Spread module', and 'deformation method' are introduced without a preceding definition or reference to prior usage; a short notational or architectural diagram would improve clarity.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. We address the single major comment point by point below, providing clarification and committing to revisions where appropriate.

read point-by-point responses

  1. Referee: [Abstract / HyperPoint Generation module] Abstract (and the description of the HyperPoint Generation module): the central motivation states that pooling operations in Transformers cause loss of local details, yet the proposed pipeline explicitly downsamples the input after offset estimation to produce the HyperPoints on which Mamba operates. No analysis is supplied showing that this downsampling step preserves the fine-grained geometry that the method claims to protect; if the downsampling discards information before or during Mamba encoding, the claimed advantage over pooling-based Transformers does not follow.

    Authors: We thank the referee for this insightful observation. The HyperPoint Generation module applies Mamba's input-dependent selection mechanism to encode features from the full input sequence prior to offset estimation and downsampling; the subsequent downsampling produces a compact set of HyperPoints that serve as the basis for the spread and deformation stages. This selection process is designed to be adaptive rather than fixed like pooling, potentially retaining more task-relevant local geometry. However, we acknowledge that the manuscript does not include dedicated analysis (e.g., ablation on downsampling ratios or quantitative measures of local feature preservation) to directly demonstrate this advantage. We will add such analysis, including visualizations and comparative metrics, to the revised manuscript. revision: yes

Circularity Check

0 steps flagged

No circularity; claims rest on external benchmarks

full rationale

The paper presents an architectural proposal (HyperPoint Generation using Mamba selection, followed by Spread and deformation modules) whose performance is asserted via quantitative/qualitative results on established point-cloud benchmarks. No equations, fitted parameters, or predictions are shown that reduce by construction to the inputs; no self-citation chains or uniqueness theorems are invoked as load-bearing; the Mamba reference is external. The derivation chain is therefore self-contained against independent data.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 2 invented entities

The central claim depends on the untested assumption that Mamba can be directly adapted to unordered point data and that the newly introduced HyperPoint modules function as described. No free parameters or external benchmarks are mentioned in the abstract.

axioms (1)
  • domain assumption Mamba's selection mechanism can effectively encode features from point cloud data
    Invoked when describing the HyperPoint Generation module that uses Mamba to process point features.
invented entities (2)
  • HyperPoints no independent evidence
    purpose: Intermediate downsampled points with offsets used for reconstruction
    New entity introduced to represent the output of the first module
  • HyperPoint Spread module no independent evidence
    purpose: Disperse HyperPoints spatially to improve coverage
    New module proposed to address concentration issues

pith-pipeline@v0.9.0 · 5741 in / 1376 out tokens · 36639 ms · 2026-05-24T02:06:47.699475+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.

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

61 extracted references · 61 canonical work pages · 6 internal anchors

  1. [1]

    Panos Achlioptas, Olga Diamanti, Ioannis Mitliagkas, and Leonidas Guibas. 2018. Learning representations and generative models for 3d point clouds. In Interna- tional conference on machine learning . PMLR, 40–49

    work page 2018

  2. [2]

    Yingjie Cai, Kwan-Yee Lin, Chao Zhang, Qiang Wang, Xiaogang Wang, and Hongsheng Li. 2022. Learning a structured latent space for unsupervised point cloud completion. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 5543–5553

    work page 2022

  3. [3]

    Zhikai Chen, Fuchen Long, Zhaofan Qiu, Ting Yao, Wengang Zhou, Jiebo Luo, and Tao Mei. 2023. AnchorFormer: Point Cloud Completion From Discriminative Nodes. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 13581–13590

    work page 2023

  4. [4]

    François Chollet. 2017. Xception: Deep learning with depthwise separable con- volutions. In Proceedings of the IEEE conference on computer vision and pattern recognition. 1251–1258

    work page 2017

  5. [5]

    Ben Fei, Weidong Yang, Wen-Ming Chen, Zhijun Li, Yikang Li, Tao Ma, Xing Hu, and Lipeng Ma. 2022. Comprehensive review of deep learning-based 3d point cloud completion processing and analysis. IEEE Transactions on Intelligent Transportation Systems 23, 12 (2022), 22862–22883

    work page 2022

  6. [6]

    Ben Fei, Weidong Yang, Wen-Ming Chen, and Lipeng Ma. 2022. VQ-DcTr: Vector- quantized autoencoder with dual-channel transformer points splitting for 3D point cloud completion. In Proceedings of the 30th ACM international conference on multimedia. 4769–4778

    work page 2022

  7. [7]

    Ben Fei, Weidong Yang, Lipeng Ma, and Wen-Ming Chen. 2023. DcTr: Noise- robust point cloud completion by dual-channel transformer with cross-attention. Pattern Recognition 133 (2023), 109051

    work page 2023

  8. [8]

    Ben Fei, Rui Zhang, Weidong Yang, Zhijun Li, and Wen-Ming Chen. 2024. Pro- gressive Growth for Point Cloud Completion by Surface-Projection Optimization. IEEE Transactions on Intelligent Vehicles (2024)

    work page 2024

  9. [9]

    Albert Gu and Tri Dao. 2023. Mamba: Linear-time sequence modeling with selective state spaces. arXiv preprint arXiv:2312.00752 (2023)

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  10. [10]

    Albert Gu, Tri Dao, Stefano Ermon, Atri Rudra, and Christopher Ré. 2020. Hippo: Recurrent memory with optimal polynomial projections. Advances in neural information processing systems 33 (2020), 1474–1487

    work page 2020

  11. [11]

    Albert Gu, Karan Goel, and Christopher Ré. 2021. Efficiently modeling long sequences with structured state spaces. arXiv preprint arXiv:2111.00396 (2021)

    work page internal anchor Pith review Pith/arXiv arXiv 2021

  12. [12]

    Albert Gu, Isys Johnson, Karan Goel, Khaled Saab, Tri Dao, Atri Rudra, and Christopher Ré. 2021. Combining recurrent, convolutional, and continuous-time models with linear state space layers. Advances in neural information processing systems 34 (2021), 572–585

    work page 2021

  13. [13]

    Tao Guo, Yinuo Wang, and Cai Meng. 2024. Mambamorph: a mamba-based backbone with contrastive feature learning for deformable mr-ct registration. arXiv preprint arXiv:2401.13934 (2024)

    work page arXiv 2024

  14. [14]

    Zitian Huang, Yikuan Yu, Jiawen Xu, Feng Ni, and Xinyi Le. 2020. Pf-net: Point fractal network for 3d point cloud completion. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 7662–7670

    work page 2020

  15. [15]

    Shanshan Li, Pan Gao, Xiaoyang Tan, and Mingqiang Wei. 2023. ProxyFormer: Proxy Alignment Assisted Point Cloud Completion with Missing Part Sensitive Transformer. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 9466–9475

    work page 2023

  16. [16]

    Dingkang Liang, Xin Zhou, Xinyu Wang, Xingkui Zhu, Wei Xu, Zhikang Zou, Xiaoqing Ye, and Xiang Bai. 2024. PointMamba: A Simple State Space Model for Point Cloud Analysis. arXiv preprint arXiv:2402.10739 (2024)

    work page arXiv 2024

  17. [17]

    Jiarun Liu, Hao Yang, Hong-Yu Zhou, Yan Xi, Lequan Yu, Yizhou Yu, Yong Liang, Guangming Shi, Shaoting Zhang, Hairong Zheng, et al . 2024. Swin- umamba: Mamba-based unet with imagenet-based pretraining. arXiv preprint arXiv:2402.03302 (2024)

    work page arXiv 2024

  18. [18]

    Yue Liu, Yunjie Tian, Yuzhong Zhao, Hongtian Yu, Lingxi Xie, Yaowei Wang, Qixiang Ye, and Yunfan Liu. 2024. Vmamba: Visual state space model. arXiv preprint arXiv:2401.10166 (2024)

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  19. [19]

    Shitong Luo and Wei Hu. 2021. Diffusion probabilistic models for 3d point cloud generation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2837–2845

    work page 2021

  20. [20]

    Zhaoyang Lyu, Zhifeng Kong, Xudong Xu, Liang Pan, and Dahua Lin. 2021. A conditional point diffusion-refinement paradigm for 3d point cloud completion. arXiv preprint arXiv:2112.03530 (2021)

    work page arXiv 2021

  21. [21]

    Jun Ma, Feifei Li, and Bo Wang. 2024. U-mamba: Enhancing long-range de- pendency for biomedical image segmentation. arXiv preprint arXiv:2401.04722 (2024)

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  22. [22]

    Harsh Mehta, Ankit Gupta, Ashok Cutkosky, and Behnam Neyshabur. 2022. Long range language modeling via gated state spaces. arXiv preprint arXiv:2206.13947 (2022)

    work page arXiv 2022

  23. [23]

    Paritosh Mittal, Yen-Chi Cheng, Maneesh Singh, and Shubham Tulsiani. 2022. Autosdf: Shape priors for 3d completion, reconstruction and generation. In Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 306–315

    work page 2022

  24. [24]

    Trung Nguyen, Quang-Hieu Pham, Tam Le, Tung Pham, Nhat Ho, and Binh- Son Hua. 2021. Point-set distances for learning representations of 3d point clouds. In Proceedings of the IEEE/CVF international conference on computer vision . 10478–10487

    work page 2021

  25. [25]

    Liang Pan. 2020. ECG: Edge-aware point cloud completion with graph convolu- tion. IEEE Robotics and Automation Letters 5, 3 (2020), 4392–4398

    work page 2020

  26. [26]

    Yatian Pang, Wenxiao Wang, Francis EH Tay, Wei Liu, Yonghong Tian, and Li Yuan. 2022. Masked autoencoders for point cloud self-supervised learning. In European conference on computer vision . Springer, 604–621

    work page 2022

  27. [27]

    Jonathan Pilault, Mahan Fathi, Orhan Firat, Chris Pal, Pierre-Luc Bacon, and Ross Goroshin. 2024. Block-state transformers. Advances in Neural Information Processing Systems 36 (2024)

    work page 2024

  28. [28]

    Maciej Pióro, Kamil Ciebiera, Krystian Król, Jan Ludziejewski, and Sebastian Jaszczur. 2024. Moe-mamba: Efficient selective state space models with mixture of experts. arXiv preprint arXiv:2401.04081 (2024)

    work page arXiv 2024

  29. [29]

    Charles R Qi, Hao Su, Kaichun Mo, and Leonidas J Guibas. 2017. Pointnet: Deep learning on point sets for 3d classification and segmentation. In Proceedings of the IEEE conference on computer vision and pattern recognition . 652–660

    work page 2017

  30. [30]

    Charles Ruizhongtai Qi, Li Yi, Hao Su, and Leonidas J Guibas. 2017. Pointnet++: Deep hierarchical feature learning on point sets in a metric space. Advances in neural information processing systems 30 (2017)

    work page 2017

  31. [31]

    Jiacheng Ruan and Suncheng Xiang. 2024. Vm-unet: Vision mamba unet for medical image segmentation. arXiv preprint arXiv:2402.02491 (2024)

    work page arXiv 2024

  32. [32]

    Jimmy TH Smith, Andrew Warrington, and Scott W Linderman. 2022. Simplified state space layers for sequence modeling. arXiv preprint arXiv:2208.04933 (2022)

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  33. [33]

    Marina Sokolova, Nathalie Japkowicz, and Stan Szpakowicz. 2006. Beyond ac- curacy, F-score and ROC: a family of discriminant measures for performance evaluation. In Australasian joint conference on artificial intelligence . Springer, 1015–1021

    work page 2006

  34. [34]

    Lyne P Tchapmi, Vineet Kosaraju, Hamid Rezatofighi, Ian Reid, and Silvio Savarese. 2019. Topnet: Structural point cloud decoder. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 383–392

    work page 2019

  35. [35]

    Keneni W Tesema, Lyndon Hill, Mark W Jones, Muneeb I Ahmad, and Gary KL Tam. 2023. Point Cloud Completion: A Survey. IEEE Transactions on Visualization and Computer Graphics (2023)

    work page 2023

  36. [36]

    Chloe Wang, Oleksii Tsepa, Jun Ma, and Bo Wang. 2024. Graph-mamba: Towards long-range graph sequence modeling with selective state spaces. arXiv preprint arXiv:2402.00789 (2024)

    work page arXiv 2024

  37. [37]

    Junxiong Wang, Tushaar Gangavarapu, Jing Nathan Yan, and Alexander M Rush. 2024. Mambabyte: Token-free selective state space model. arXiv preprint arXiv:2401.13660 (2024)

    work page arXiv 2024

  38. [38]

    Xiaogang Wang, Marcelo H Ang Jr, and Gim Hee Lee. 2020. Cascaded refinement network for point cloud completion. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 790–799

    work page 2020

  39. [39]

    Yida Wang, David Joseph Tan, Nassir Navab, and Federico Tombari. 2022. Learn- ing local displacements for point cloud completion. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition . 1568–1577

    work page 2022

  40. [40]

    Weikun Wu, Yan Zhang, David Wang, and Yunqi Lei. 2020. SK-Net: Deep learning on point cloud via end-to-end discovery of spatial keypoints. In Proceedings of the AAAI Conference on Artificial Intelligence , Vol. 34. 6422–6429

    work page 2020

  41. [41]

    Yaqi Xia, Yan Xia, Wei Li, Rui Song, Kailang Cao, and Uwe Stilla. 2021. Asfm- net: Asymmetrical siamese feature matching network for point completion. In Proceedings of the 29th ACM international conference on multimedia . 1938–1947

    work page 2021

  42. [42]

    Peng Xiang, Xin Wen, Yu-Shen Liu, Yan-Pei Cao, Pengfei Wan, Wen Zheng, and Zhizhong Han. 2021. Snowflakenet: Point cloud completion by snowflake point deconvolution with skip-transformer. InProceedings of the IEEE/CVF international conference on computer vision . 5499–5509

    work page 2021

  43. [43]

    Haozhe Xie, Hongxun Yao, Shangchen Zhou, Jiageng Mao, Shengping Zhang, and Wenxiu Sun. 2020. Grnet: Gridding residual network for dense point cloud completion. In European Conference on Computer Vision . Springer, 365–381

    work page 2020

  44. [44]

    Zhaohu Xing, Tian Ye, Yijun Yang, Guang Liu, and Lei Zhu. 2024. Segmamba: Long-range sequential modeling mamba for 3d medical image segmentation. arXiv preprint arXiv:2401.13560 (2024). 3DMambaComplete, School of Computer Science, Fudan University Yixuan Li, Weidong Yang, and Ben Fei

    work page arXiv 2024

  45. [45]

    Xingguang Yan, Liqiang Lin, Niloy J Mitra, Dani Lischinski, Daniel Cohen-Or, and Hui Huang. 2022. Shapeformer: Transformer-based shape completion via sparse representation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 6239–6249

    work page 2022

  46. [46]

    Yaoqing Yang, Chen Feng, Yiru Shen, and Dong Tian. 2018. Foldingnet: Point cloud auto-encoder via deep grid deformation. In Proceedings of the IEEE confer- ence on computer vision and pattern recognition . 206–215

    work page 2018

  47. [47]

    Yijun Yang, Zhaohu Xing, and Lei Zhu. 2024. Vivim: a video vision mamba for medical video object segmentation. arXiv preprint arXiv:2401.14168 (2024)

    work page arXiv 2024

  48. [48]

    Zi Ye and Tianxiang Chen. 2024. P-Mamba: Marrying Perona Malik Diffusion with Mamba for Efficient Pediatric Echocardiographic Left Ventricular Segmentation. arXiv preprint arXiv:2402.08506 (2024)

    work page arXiv 2024

  49. [49]

    Xumin Yu, Yongming Rao, Ziyi Wang, Zuyan Liu, Jiwen Lu, and Jie Zhou. 2021. Pointr: Diverse point cloud completion with geometry-aware transformers. In Proceedings of the IEEE/CVF international conference on computer vision . 12498– 12507

    work page 2021

  50. [50]

    Xumin Yu, Lulu Tang, Yongming Rao, Tiejun Huang, Jie Zhou, and Jiwen Lu

    work page

  51. [51]

    In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Point-bert: Pre-training 3d point cloud transformers with masked point modeling. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 19313–19322

    work page

  52. [52]

    Wentao Yuan, Tejas Khot, David Held, Christoph Mertz, and Martial Hebert. 2018. Pcn: Point completion network. In 2018 international conference on 3D vision (3DV). IEEE, 728–737

    work page 2018

  53. [53]

    Junzhe Zhang, Xinyi Chen, Zhongang Cai, Liang Pan, Haiyu Zhao, Shuai Yi, Chai Kiat Yeo, Bo Dai, and Chen Change Loy. 2021. Unsupervised 3d shape completion through gan inversion. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 1768–1777

    work page 2021

  54. [54]

    Rui Zhang, Jingyi Xu, Weidong Yang, Lipeng Ma, Menglong Chen, and Ben Fei

    work page

  55. [55]

    In ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

    Learning Density Regulated and Multi-View Consistent Unsigned Distance Fields. In ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 8366–8370

    work page 2024

  56. [56]

    Tao Zhang, Xiangtai Li, Haobo Yuan, Shunping Ji, and Shuicheng Yan. 2024. Point Could Mamba: Point Cloud Learning via State Space Model. arXiv preprint arXiv:2403.00762 (2024)

    work page arXiv 2024

  57. [57]

    Xuancheng Zhang, Yutong Feng, Siqi Li, Changqing Zou, Hai Wan, Xibin Zhao, Yandong Guo, and Yue Gao. 2021. View-guided point cloud completion. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 15890–15899

    work page 2021

  58. [58]

    Zhuoran Zheng and Jun Zhang. 2024. FD-Vision Mamba for Endoscopic Exposure Correction. arXiv preprint arXiv:2402.06378 (2024)

    work page arXiv 2024

  59. [59]

    Haoran Zhou, Yun Cao, Wenqing Chu, Junwei Zhu, Tong Lu, Ying Tai, and Chengjie Wang. 2022. Seedformer: Patch seeds based point cloud completion with upsample transformer. In European conference on computer vision . Springer, 416–432

    work page 2022

  60. [60]

    Linqi Zhou, Yilun Du, and Jiajun Wu. 2021. 3d shape generation and comple- tion through point-voxel diffusion. In Proceedings of the IEEE/CVF international conference on computer vision . 5826–5835

    work page 2021

  61. [61]

    Lianghui Zhu, Bencheng Liao, Qian Zhang, Xinlong Wang, Wenyu Liu, and Xinggang Wang. 2024. Vision mamba: Efficient visual representation learning with bidirectional state space model. arXiv preprint arXiv:2401.09417 (2024)

    work page internal anchor Pith review Pith/arXiv arXiv 2024