Pith. sign in

REVIEW 2 major objections 8 minor 64 references

Part Segmentation and Motion Estimation for Articulated Objects with Dynamic 3D Gaussians

T0 review · 2 major / 8 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Dynamic 3D Gaussians with time-shared scales recover part segmentation, motion, and kinematic trees from point clouds with no correspondences, improving occlusion robustness by 13%.

desk verdict Solid per-part Gaussian representation for joint segmentation and kinematic estimation from unregistered point clouds, but the occlusion robustness claim needs one more ablation to separate representation from kinematic fine-tuning. read the letter →

arxiv 2506.22718 v2 pith:IX6PLQU2 submitted 2025-06-28 cs.CV

Jun-Jee Chao , Qingyuan Jiang , Volkan Isler This is my paper · ORCID
classification cs.CV
keywords articulatedobjectspartsegmentationmotionestimation3DGaussianspointcloudskinematictreeocclusionrobustnessre-articulation
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

The paper claims that a sequence of 3D point clouds of an articulated object, even when no point matches exist across frames, can be turned into part segmentation, per-part motions, and a kinematic tree by representing each rigid part as a dynamic 3D Gaussian whose pose changes over time while its shape scale stays fixed. This matters because real captures from multiple asynchronous sensors or with viewpoint occlusion break methods that require tracking the same point over time. The paper reports that on standard benchmarks this point-distribution approach matches or beats correspondence-based methods, and on occluded point clouds its part segmentation improves on the leading baseline by 13% in Rand index.

What carries the argument

The central object is a dynamic 3D Gaussian per rigid part: covariance $\Sigma = R S (R S)^T$, center $\mu$, and pose $T \in SE(3)$, with $K$ time-varying rotations and translations and one time-shared scale $s$ per part. Assignment uses squared Mahalanobis distance with Gumbel-Softmax differentiability; transformation $H^{k \to t}$ moves assigned points to other time steps via the part poses, allowing fusion of all frames. The kinematic tree is estimated by a minimum spanning tree over pairwise spatial distance and 1-DOF relative-motion error, then joint parameters are refined through forward kinematics.

What would settle it

Run the method on a point-cloud sequence of an articulated object whose ground-truth parts include one hollow or strongly elongated rigid part, such as a hinged box lid or a bent robot link, and check whether points on the concave side or inside the hollow are consistently labeled as a different part; consistent mislabeling of exactly those points, while the rest of the object segments correctly, would confirm that the single-Gaussian, shared-scale representation is the limiting factor.

Watch

Extended reading notes

Core claim

The paper's central claim is that a rigid articulated part can be identified with a single 3D Gaussian whose rotation and center move with the part while its scale stays fixed over time. Given a sequence of arbitrarily sampled point clouds, optimizing these Gaussians with a maximum-likelihood objective plus separation, Chamfer, earth-mover, and flow terms yields per-frame part labels by smallest Mahalanobis distance, per-part poses from the Gaussian transformations, and a kinematic tree from spatial proximity and 1-DOF relative motion. On RoboArt the paper reports Rand index 0.90 against 0.86 for the single-canonical-frame baseline, and on its occlusion-hardened Occluded-RoboArt variant it reports Rand index 0.85 against 0.75, a 13% relative gain. The paper also shows the estimated model can re-articulate the object to unseen poses.

Load-bearing premise

The load-bearing premise is that every rigid part's point distribution can be captured by a single 3D Gaussian whose scale is constant across time; parts that are elongated, L-shaped, or hollow violate this, and boundary points would then be assigned to the wrong part.

Editorial extensions

If this is right

  • Part labels can be produced for every observed frame from a single fitted model, removing the need to choose a canonical frame where all parts must be visible.
  • Points that are sampled in only one frame can still be transported to other frames by following their assigned Gaussian's pose, so occluded frames do not destroy the motion estimate.
  • The model outputs a kinematic tree with joint parameters, so a fitted object can be re-articulated to novel poses, not just reconstructed at observed times.
  • The method does not require the same point set to persist across frames, making asynchronous multi-sensor point-cloud sequences a workable input setting.

Reading between the lines

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

  • A testable extension the paper does not run is to feed it an articulated object with an elongated, curved, or hollow rigid part; the shared-scale single-Gaussian assumption predicts boundary points at the concave side or hollow region will be mislabeled, and if so a mixture or non-Gaussian shape model would be the natural repair.
  • Because the optimization uses only point coordinates, the same representation could be lifted to RGB-D or multi-view video by first converting frames to point clouds; the paper demonstrates only direct point-cloud sequences.
  • The method's need to rerun with several $m$ values to select the part count is a practical bottleneck; a learned or adaptive estimate of $m$, which the authors flag as future work, would make the method a drop-in component for articulation discovery.
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

2 major / 8 minor

Summary. The paper presents a three-stage optimization method for articulated object analysis from a sequence of arbitrarily sampled 3D point clouds. In Stage 1, m dynamic 3D Gaussians (time-varying rotation and translation, time-shared scale) are fitted with a weighted combination of Gaussian maximum-likelihood, separation, Chamfer, earth-mover, and scene-flow losses; segmentation is obtained by Mahalanobis assignment, and point transformations are obtained by composing Gaussian poses across time. Stage 2 estimates a kinematic tree from pairwise spatial proximity and 1-DOF motion criteria, and Stage 3 fine-tunes joint parameters under kinematic constraints. Experiments on RoboArt, Sapien, and two occlusion-augmented RoboArt datasets report improved reconstruction, flow, segmentation Rand Index, tree edit distance, and reanimation error compared with MultiBodySync, WatchItMove, and Reart, including a 13% Rand Index gain on occluded data.

Significance. If the claims are fully supported, the paper would provide a simple, category-agnostic alternative to correspondence-based methods that can also handle partial observations, which is practically useful for robotics and 4D reconstruction. The paper's strengths are the compact parametric representation, the joint optimization of segmentation and motion, and the ablation study showing that every loss term contributes on the validation set. However, the headline occlusion-robustness result is currently under-supported because the loss design leaves fully occluded parts unconstrained in Stage 1, and the comparison protocol for the 13% gain is not fully specified.

major comments (2)
  1. [Section 4.4 and Eqs. (3)-(7)] The occlusion-robustness claim is not supported by the reported experiments. For a time step t in which a part is completely invisible, no point in X^t is generated by that part, so L_MLE in Eq. (3) provides essentially no gradient for that part's R^t and mu^t. The one-directional modification of L_CD and L_EMD requires each observed point in X^t to have a nearby point in the fused cloud, but it never penalizes fused points from the occluded part, because the matching is not symmetric. The flow loss in Eq. (7) is also uninformative for points absent from both X^t and X^{t+1}. Hence Stage 1 leaves the pose of a fully occluded part underdetermined, and only the kinematic fine-tuning in Section 3.3 can determine it. Table 3 reports large gains on Occluded-RoboArt but contains no ablation that separates Stage 1 from Stage 3, and no pose error is reported specifically for parts that disappear. Without such an analysis, the abstract's claim that the method is robust to completely occluded parts cannot be attributed to the dynamic-Gaussian representation rather than to kinematic extrapolation.
  2. [Table 3 and abstract] The comparison protocol behind the '13%' claim is not apples-to-apples. Table 3 reports 'Cano Frame RI' for Reart and both 'Cano Frame RI' and 'Mean RI' for Ours, but Reart evaluates segmentation only on a single canonical frame selected by its own optimization, whereas Ours evaluates either a canonical frame or the mean over all frames; the text is ambiguous about which value is used for the 13% comparison. If Reart's best single canonical frame is compared against Ours's mean over all frames, the comparison favors Ours. The paper should report Reart's mean RI over all frames (or Ours's canonical-frame RI under the same protocol) and should state how the canonical frame is chosen for Ours. Without this, the headline improvement is uninterpretable.
minor comments (8)
  1. [Section 4.1] Please report variances or multiple runs for the main tables, because optimization-based methods can be sensitive to initialization and the reported differences in Table 2 are small.
  2. [Table 3] The row labels should specify which metric is computed for Reart versus Ours, and whether 'Mean RI' exists for Reart; the canonical-frame selection procedure for Ours should also be stated.
  3. [Eq. (7)] Define \hat X^t before using it in the flow loss; Eq. (2) defines H^{k\to t} but not the notation \hat X^t.
  4. [Section 3.1 and Section 3.2] Clarify the relationship between the m Gaussians and the final rigid parts, since Section 3.2 merges Gaussians; the text sometimes says 'each Gaussian represents a single rigid part' and sometimes suggests over-segmentation with later merging.
  5. [Table 1] The 'Ours†' row is useful, but the table caption should state explicitly that it is without kinematic fine-tuning; the main text does mention this, but a caption would improve clarity.
  6. [Section 4.5] Table 4's 'Ours' row reports Recons Error 1.26 while Table 1's RoboArt test 'Ours' reports 0.88; state clearly that these are validation and test sets, respectively.
  7. [References and text] Reference [14] contains a typo: 'Gumble-Softmax' should be 'Gumbel-Softmax'; the conclusion also contains a duplicated 'for' in 'for for many real-world objects.'
  8. [Section 3.1 and Limitations] The selection of m by minimizing L_CD after convergence is a possible source of selection bias, since L_CD is also part of the optimization objective; a short discussion of this risk in the limitations paragraph would be helpful.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the dynamic-Gaussian optimization is evaluated against external benchmarks, and no equation reduces to its own inputs; only minor non-load-bearing self-citations appear.

full rationale

The paper's central derivation is a self-contained optimization. Segmentation is defined by Mahalanobis assignment in Eq. 1, Gaussian parameters are optimized via the maximum-likelihood loss in Eq. 3, and cross-time consistency is enforced by Chamfer, Earth-mover, and flow losses in Eqs. 5-7. These objectives are all functions of the observed point clouds and the optimized Gaussian parameters; none of them encodes the ground-truth part labels or the final benchmark metrics. The reported Rand Index, Reconstruction Error, Flow Error, and Tree Edit Distance are measured against external public benchmarks (RoboArt and Sapien), so they are not equal to any fitted constant by construction. The selection of the number of parts m by minimizing the final training loss L_CD is standard model selection on the objective, not a prediction of a closely related quantity; Table 2's 'Ours*' is explicitly labeled an upper bound. The only self-citations, [3] and [8], appear in a related-work list of category-based methods and are not load-bearing for the method's validity. The occlusion-robustness concern raised by the skeptical reader is an evidentiary gap, not a circularity: for fully occluded parts, some loss terms may indeed be uninformative before kinematic fine-tuning, and the paper's own Limitations section concedes failure when large portions are unseen, but no equation in the paper reduces to its inputs. Therefore no circular step is identified.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The method relies on user-chosen part counts and loss weights, a single-Gaussian-per-part shape assumption, a tree-like 1-DOF kinematic assumption, and an external pretrained scene flow network. No new physical entities are proposed.

free parameters (3)
  • number of Gaussians m = 10-15 on RoboArt, 3/5/7 on Sapien, 20 on occlusion sets
    User supplied; selected by running multiple optimizations and picking the lowest L_CD. It directly determines segmentation granularity and is the main hyperparameter.
  • loss weights lambda_MLE, lambda_sep, lambda_CD, lambda_EMD, lambda_flow = not stated in main text
    The five weighted loss terms in the total loss require these weights; without them the exact objective is not reproducible from the paper alone.
  • separation margin alpha in L_sep = not stated in main text
    Controls the minimum allowed Mahalanobis distance between Gaussian centers; affects how finely parts can be separated.
assumptions (4)
  • domain assumption Each rigid part's point distribution is a single 3D Gaussian with a scale shared across all time steps.
    Section 3.1: 'each Gaussian represents the point distribution and motion of a single rigid part' and 'sharing the same s across all time steps'.
  • domain assumption The kinematic structure is acyclic and every parent-child pair exhibits 1-DOF relative motion.
    Section 5 Limitations: 'our method assumes that the kinematic structure is acyclic and the parent-child part pairs exhibit 1-DOF relative motion.'
  • domain assumption The pretrained scene flow network g supplies reliable point-wise flow for arbitrary and partially occluded point clouds.
    Section 3.1 Flow loss uses g from [12,26]; the method's motion cue depends on this external network, and its behavior under arbitrary sampling is not analyzed.
  • ad hoc to paper Equal Gaussian weights are enforced so each Gaussian maps to one rigid part.
    Section 3.1: 'we enforce equal weights for all Gaussians to ensure that each Gaussian represents a single rigid part.' This is a modeling choice specific to this paper, not a standard GMM constraint.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Part Segmentation and Motion Estimation for Articulated Objects with Dynamic 3D Gaussians." pith.science (2026). https://pith.science/paper/IX6PLQU2

@misc{pith2026250622718,
  author       = {Pith},
  title        = {Pith review of: Part Segmentation and Motion Estimation for Articulated Objects with Dynamic 3D Gaussians},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IX6PLQU2}},
  note         = {Machine review of arXiv:2506.22718}
}
read the original abstract

Part segmentation and motion estimation are two fundamental problems for articulated object motion analysis. In this paper, we present a method to solve these two problems jointly from a sequence of observed point clouds of a single articulated object. The main challenge in our problem setting is that the point clouds are not assumed to be generated by a fixed set of moving points. Instead, each point cloud in the sequence could be an arbitrary sampling of the object surface at that particular time step. Such scenarios occur when the object undergoes major occlusions, or if the dataset is collected using measurements from multiple sensors asynchronously. In these scenarios, methods that rely on tracking point correspondences are not appropriate. We present an alternative approach based on a compact but effective representation where we represent the object as a collection of simple building blocks modeled as 3D Gaussians. We parameterize the Gaussians with time-dependent rotations, translations, and scales that are shared across all time steps. With our representation, part segmentation can be achieved by building correspondences between the observed points and the Gaussians. Moreover, the transformation of each point across time can be obtained by following the poses of the assigned Gaussian (even when the point is not observed). Experiments show that our method outperforms existing methods that solely rely on finding point correspondences. Additionally, we extend existing datasets to emulate real-world scenarios by considering viewpoint occlusions. We further demonstrate that our method is more robust to missing points as compared to existing approaches on these challenging datasets, even when some parts are completely occluded in some time-steps. Notably, our part segmentation performance outperforms the state-of-the-art method by 13% on point clouds with occlusions.

Figures

Figures reproduced from arXiv: 2506.22718 by the authors.

Figure 1
Figure 1. Problem formulation: (a) Given a dynamic point cloud sequence of an articulated [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Our optimization pipeline. In each iteration, we sample a point cloud from [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Qualitative results on the RoboArt dataset. (i) The third row shows our estimated [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: Qualitative results on Occluded￾RoboArt. Our method separates the parts where the baseline fails. 4.4 Result on partial point clouds We compare our method with Reart [26], which models part segmentation with a coordinate￾based MLP, on two challenging datasets: Partial-…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

64 extracted references · 54 canonical work pages

  1. [1]

    Learning to generalize kinematic models to novel objects

    Ben Abbatematteo, Stefanie Tellex, and George Konidaris. Learning to generalize kinematic models to novel objects. InProceedings of the 3rd Conference on Robot Learning, 2019

    work page 2019

  2. [2]

    Learning to infer kinematic hierarchies for novel object instances

    Hameed Abdul-Rashid, Miles Freeman, Ben Abbatematteo, George Konidaris, and Daniel Ritchie. Learning to infer kinematic hierarchies for novel object instances. In 2022 International Conference on Robotics and Automation (ICRA), pages 8461–8467. IEEE, 2022

    work page 2022

  3. [3]

    Category-level global camera pose estimation with multi-hypothesis point cloud correspondences

    Jun-Jee Chao, Selim Engin, Nicolai Häni, and V olkan Isler. Category-level global camera pose estimation with multi-hypothesis point cloud correspondences. In2023 IEEE International Conference on Robotics and Automation (ICRA), pages 3800–3807. IEEE, 2023

    work page 2023

  4. [4]

    A benchmark for 3d mesh segmentation.Acm transactions on graphics (tog), 28(3):1–12, 2009

    Xiaobai Chen, Aleksey Golovinskiy, and Thomas Funkhouser. A benchmark for 3d mesh segmentation.Acm transactions on graphics (tog), 28(3):1–12, 2009

    work page 2009

  5. [5]

    MIT press, 2022

    Thomas H Cormen, Charles E Leiserson, Ronald L Rivest, and Clifford Stein.Intro- duction to algorithms. MIT press, 2022

    work page 2022

  6. [6]

    On implementing 2d rectangular assignment algorithms.IEEE Trans- actions on Aerospace and Electronic Systems, 52(4):1679–1696, 2016

    David F Crouse. On implementing 2d rectangular assignment algorithms.IEEE Trans- actions on Aerospace and Electronic Systems, 52(4):1679–1696, 2016

    work page 2016

  7. [7]

    Banana: Banach fixed-point network for pointcloud segmentation with inter-part equiv- ariance.Advances in Neural Information Processing Systems, 36, 2024

    Congyue Deng, Jiahui Lei, William B Shen, Kostas Daniilidis, and Leonidas J Guibas. Banana: Banach fixed-point network for pointcloud segmentation with inter-part equiv- ariance.Advances in Neural Information Processing Systems, 36, 2024

    work page 2024

  8. [8]

    3D Surface Reconstruction in the Wild by Deforming Shape Priors from Synthetic Data

    Nicolai Häni, Jun-Jee Chao, and V olkan Isler. 3d surface reconstruction in the wild by deforming shape priors from synthetic data.arXiv preprint arXiv:2302.12883, 2023

    work page Pith review arXiv 2023

Show all 64 references
  1. [9]

    Nonparametric object and parts modeling with lie group dynamics

    David S Hayden, Jason Pacheco, and John W Fisher. Nonparametric object and parts modeling with lie group dynamics. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 7426–7435, 2020

    2020

  2. [10]

    Carto: Cate- gory and joint agnostic reconstruction of articulated objects

    Nick Heppert, Muhammad Zubair Irshad, Sergey Zakharov, Katherine Liu, Rares An- drei Ambrus, Jeannette Bohg, Abhinav Valada, and Thomas Kollar. Carto: Cate- gory and joint agnostic reconstruction of articulated objects. InProceedings of the IEEE/CVF Conference on Computer Visi...

    2023

  3. [11]

    Ditto in the house: Building articu- lation models of indoor scenes through interactive perception

    Cheng-Chun Hsu, Zhenyu Jiang, and Yuke Zhu. Ditto in the house: Building articu- lation models of indoor scenes through interactive perception. InIEEE International Conference on Robotics and Automation (ICRA), 2023

    2023

  4. [12]

    Multibodysync: Multi-body segmentation and motion estimation via 3d scan synchronization

    Jiahui Huang, He Wang, Tolga Birdal, Minhyuk Sung, Federica Arrigoni, Shi-Min Hu, and Leonidas J Guibas. Multibodysync: Multi-body segmentation and motion estimation via 3d scan synchronization. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognitio...

    2021

  5. [13]

    Screwnet: Category- independent articulation model estimation from depth images using screw theory

    Ajinkya Jain, Rudolf Lioutikov, Caleb Chuck, and Scott Niekum. Screwnet: Category- independent articulation model estimation from depth images using screw theory. In2021 IEEE International Conference on Robotics and Automation (ICRA), pages 13670–13677. IEEE, 2021

    2021

  6. [14]

    Categorical reparametrization with gumble- softmax

    Eric Jang, Shixiang Gu, and Ben Poole. Categorical reparametrization with gumble- softmax. InInternational Conference on Learning Representations (ICLR 2017), 2017

    2017

  7. [15]

    Ditto: Building digital twins of ar- ticulated objects from interaction

    Zhenyu Jiang, Cheng-Chun Hsu, and Yuke Zhu. Ditto: Building digital twins of ar- ticulated objects from interaction. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 5616–5626, 2022

    2022

  8. [16]

    A shortest augmenting path algorithm for dense and sparse linear assignment problems.Computing (Wien

    R JONKER and A VOLGENANT. A shortest augmenting path algorithm for dense and sparse linear assignment problems.Computing (Wien. Print), 38(4):325–340, 1987

    1987

  9. [17]

    Direct visibility of point sets

    Sagi Katz, Ayellet Tal, and Ronen Basri. Direct visibility of point sets. InACM SIG- GRAPH 2007 papers, pages 24–es. 2007

    2007

  10. [18]

    3d gaussian splatting for real-time radiance field rendering.ACM Transactions on Graphics, 42(4), July 2023

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimkühler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering.ACM Transactions on Graphics, 42(4), July 2023. URLhttps://repo-sam.inria.fr/fungraph/ 3d-gaussian-splatting/

    2023

  11. [19]

    Camm: Building category-agnostic and animatable 3d models from monocular videos

    Tianshu Kuai, Akash Karthikeyan, Yash Kant, Ashkan Mirzaei, and Igor Gilitschen- ski. Camm: Building category-agnostic and animatable 3d models from monocular videos. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, pages 65...

    2023

  12. [20]

    Gart: Gaussian articulated template models

    Jiahui Lei, Yufu Wang, Georgios Pavlakos, Lingjie Liu, and Kostas Daniilidis. Gart: Gaussian articulated template models. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 19876–19887, 2024

    2024

  13. [21]

    Category-level articulated object pose estimation

    Xiaolong Li, He Wang, Li Yi, Leonidas J Guibas, A Lynn Abbott, and Shuran Song. Category-level articulated object pose estimation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 3706–3715, 2020

    2020

  14. [22]

    Lepard: Learning explicit part discovery for 3d articulated shape re- construction

    Di Liu, Anastasis Stathopoulos, Qilong Zhangli, Yunhe Gao, and Dimitris Metaxas. Lepard: Learning explicit part discovery for 3d articulated shape re- construction. In A. Oh, T. Naumann, A. Globerson, K. Saenko, M. Hardt, and S. Levine, editors,Advances in Neural Information P...

    2023

  15. [23]

    Semi-weakly supervised object kinematic motion prediction

    Gengxin Liu, Qian Sun, Haibin Huang, Chongyang Ma, Yulan Guo, Li Yi, Hui Huang, and Ruizhen Hu. Semi-weakly supervised object kinematic motion prediction. InPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023

    2023

  16. [24]

    Paris: Part-level reconstruction and motion analysis for articulated objects

    Jiayi Liu, Ali Mahdavi-Amiri, and Manolis Savva. Paris: Part-level reconstruction and motion analysis for articulated objects. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 352–363, 2023. CHAO ET AL.: ARTICULA TED OBJECT PART SEGMENTA TION13

    2023

  17. [25]

    Partslip: Low-shot part segmentation for 3d point clouds via pretrained image- language models

    Minghua Liu, Yinhao Zhu, Hong Cai, Shizhong Han, Zhan Ling, Fatih Porikli, and Hao Su. Partslip: Low-shot part segmentation for 3d point clouds via pretrained image- language models. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 217...

    2023

  18. [26]

    Building rearticulable models for arbitrary 3d objects from 4d point clouds

    Shaowei Liu, Saurabh Gupta, and Shenlong Wang. Building rearticulable models for arbitrary 3d objects from 4d point clouds. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 21138–21147, 2023

    2023

  19. [27]

    Meteornet: Deep learning on dy- namic 3d point cloud sequences

    Xingyu Liu, Mengyuan Yan, and Jeannette Bohg. Meteornet: Deep learning on dy- namic 3d point cloud sequences. InProceedings of the IEEE/CVF International Con- ference on Computer Vision, pages 9246–9255, 2019

    2019

  20. [28]

    Self-supervised category-level articulated object pose estimation with part-level se (3) equivariance

    Xueyi Liu, Ji Zhang, Ruizhen Hu, Haibin Huang, He Wang, and Li Yi. Self-supervised category-level articulated object pose estimation with part-level se (3) equivariance. In The Eleventh International Conference on Learning Representations, 2023

    2023

  21. [29]

    Relation-shape con- volutional neural network for point cloud analysis

    Yongcheng Liu, Bin Fan, Shiming Xiang, and Chunhong Pan. Relation-shape con- volutional neural network for point cloud analysis. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8895–8904, 2019

    2019

  22. [30]

    Matthew Loper, Naureen Mahmood, Javier Romero, Gerard Pons-Moll, and Michael J. Black. SMPL: A skinned multi-person linear model.ACM Trans. Graphics (Proc. SIGGRAPH Asia), 34(6):248:1–248:16, October 2015

    2015

  23. [31]

    Dynamic 3d gaussians: Tracking by persistent dynamic view synthesis

    Jonathon Luiten, Georgios Kopanas, Bastian Leibe, and Deva Ramanan. Dynamic 3d gaussians: Tracking by persistent dynamic view synthesis. In2024 International Conference on 3D Vision (3DV), pages 800–809. IEEE, 2024

    2024

  24. [32]

    Modern robotics: Mechanics, planning, and control

    Kevin M Lynch and Frank C Park. Modern robotics: Mechanics, planning, and control. 2017

    2017

  25. [33]

    The concrete distribution: A continuous relaxation of discrete random variables

    C Maddison, A Mnih, and Y Teh. The concrete distribution: A continuous relaxation of discrete random variables. InProceedings of the international conference on learning Representations, 2017

    2017

  26. [34]

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

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

    1996

  27. [35]

    Structure from action: Learning interactions for 3d articulated object structure discovery

    Neil Nie, Samir Yitzhak Gadre, Kiana Ehsani, and Shuran Song. Structure from action: Learning interactions for 3d articulated object structure discovery. In2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 1222–1229. IEEE, 2023

    2023

  28. [36]

    Watch it move: Unsupervised discovery of 3d joints for re-posing of articulated ob- jects

    Atsuhiro Noguchi, Umar Iqbal, Jonathan Tremblay, Tatsuya Harada, and Orazio Gallo. Watch it move: Unsupervised discovery of 3d joints for re-posing of articulated ob- jects. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 3677–3687, 2022

    2022

  29. [37]

    Efficient computation of the tree edit distance

    Mateusz Pawlik and Nikolaus Augsten. Efficient computation of the tree edit distance. ACM Transactions on Database Systems (TODS), 40(1):1–40, 2015. 14CHAO ET AL.: ARTICULA TED OBJECT PART SEGMENTA TION

    2015

  30. [38]

    Pointnet++: Deep hier- archical feature learning on point sets in a metric space.Advances in neural information processing systems, 30, 2017

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

    2017

  31. [39]

    Pointnext: Revisiting pointnet++ with improved training and scaling strategies.Advances in neural information processing systems, 35:23192–23204, 2022

    Guocheng Qian, Yuchen Li, Houwen Peng, Jinjie Mai, Hasan Hammoud, Mohamed Elhoseiny, and Bernard Ghanem. Pointnext: Revisiting pointnet++ with improved training and scaling strategies.Advances in neural information processing systems, 35:23192–23204, 2022

    2022

  32. [40]

    Gaussian mixture models.Encyclopedia of biometrics, 741 (659-663):3, 2009

    Douglas A Reynolds et al. Gaussian mixture models.Encyclopedia of biometrics, 741 (659-663):3, 2009

    2009

  33. [41]

    Javier Romero, Dimitrios Tzionas, and Michael J. Black. Embodied hands: Model- ing and capturing hands and bodies together.ACM Transactions on Graphics, (Proc. SIGGRAPH Asia), 36(6), November 2017

    2017

  34. [42]

    Self-supervised learning of part mobility from point cloud sequence

    Yahao Shi, Xinyu Cao, and Bin Zhou. Self-supervised learning of part mobility from point cloud sequence. InComputer Graphics F orum, volume 40, pages 104–116. Wiley Online Library, 2021

    2021

  35. [43]

    Re- acto: Reconstructing articulated objects from a single video

    Chaoyue Song, Jiacheng Wei, Chuan Sheng Foo, Guosheng Lin, and Fayao Liu. Re- acto: Reconstructing articulated objects from a single video. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 5384– 5395, 2024

    2024

  36. [44]

    Performance of bayesian model selection criteria for gaussian mixture models.Frontiers of statistical decision making and bayesian analysis, 2:113–130, 2010

    Russell J Steele and Adrian E Raftery. Performance of bayesian model selection criteria for gaussian mixture models.Frontiers of statistical decision making and bayesian analysis, 2:113–130, 2010

    2010

  37. [45]

    Geometry and screw theory for robotics

    Stefano Stramigioli and Herman Bruyninckx. Geometry and screw theory for robotics. Tutorial during ICRA, 2001:75, 2001

    2001

  38. [46]

    Kpconv: Flexible and deformable con- volution for point clouds

    Hugues Thomas, Charles R Qi, Jean-Emmanuel Deschaud, Beatriz Marcotegui, François Goulette, and Leonidas J Guibas. Kpconv: Flexible and deformable con- volution for point clouds. InProceedings of the IEEE/CVF international conference on computer vision, pages 6411–6420, 2019

    2019

  39. [47]

    Cla-nerf: Category- level articulated neural radiance field

    Wei-Cheng Tseng, Hung-Ju Liao, Lin Yen-Chen, and Min Sun. Cla-nerf: Category- level articulated neural radiance field. In2022 International Conference on Robotics and Automation (ICRA), pages 8454–8460. IEEE, 2022

    2022

  40. [48]

    Neus: Learning neural implicit surfaces by volume rendering for multi-view reconstruction.Advances in Neural Information Processing Systems, 34:27171–27183, 2021

    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.Advances in Neural Information Processing Systems, 34:27171–27183, 2021

    2021

  41. [49]

    AdaAfford: Learning to adapt manipulation affordance for 3d articulated objects via few-shot interactions

    Yian Wang, Ruihai Wu, Kaichun Mo, Jiaqi Ke, Qingnan Fan, Leonidas Guibas, and Hao Dong. AdaAfford: Learning to adapt manipulation affordance for 3d articulated objects via few-shot interactions. 2022

    2022

  42. [50]

    Dynamic graph cnn for learning on point clouds.ACM Trans- actions on Graphics (tog), 38(5):1–12, 2019

    Yue Wang, Yongbin Sun, Ziwei Liu, Sanjay E Sarma, Michael M Bronstein, and Justin M Solomon. Dynamic graph cnn for learning on point clouds.ACM Trans- actions on Graphics (tog), 38(5):1–12, 2019. CHAO ET AL.: ARTICULA TED OBJECT PART SEGMENTA TION15

    2019

  43. [51]

    Self- supervised neural articulated shape and appearance models

    Fangyin Wei, Rohan Chabra, Lingni Ma, Christoph Lassner, Michael Zollhoefer, Szy- mon Rusinkiewicz, Chris Sweeney, Richard Newcombe, and Mira Slavcheva. Self- supervised neural articulated shape and appearance models. InProceedings IEEE/CVF Conference on Computer Vision and Pa...

    2022

  44. [52]

    Magicpony: Learning articulated 3d animals in the wild

    Shangzhe Wu, Ruining Li, Tomas Jakab, Christian Rupprecht, and Andrea Vedaldi. Magicpony: Learning articulated 3d animals in the wild. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8792– 8802, 2023

    2023

  45. [53]

    Pointpwc-net: Cost volume on point clouds for (self-) supervised scene flow estimation

    Wenxuan Wu, Zhi Yuan Wang, Zhuwen Li, Wei Liu, and Li Fuxin. Pointpwc-net: Cost volume on point clouds for (self-) supervised scene flow estimation. InEuropean Conference on Computer Vision, pages 88–107, 2020

    2020

  46. [54]

    CASA: Category-agnostic skeletal animal reconstruction

    Yuefan Wu*, Zeyuan Chen*, Shaowei Liu, Zhongzheng Ren, and Shenlong Wang. CASA: Category-agnostic skeletal animal reconstruction. InNeural Information Pro- cessing Systems (NeurIPS), 2022

    2022

  47. [55]

    Sapien: A simulated part-based inter- active environment

    Fanbo Xiang, Yuzhe Qin, Kaichun Mo, Yikuan Xia, Hao Zhu, Fangchen Liu, Minghua Liu, Hanxiao Jiang, Yifu Yuan, He Wang, et al. Sapien: A simulated part-based inter- active environment. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 11...

    2020

  48. [56]

    Lasr: Learning articulated shape reconstruction from a monocular video

    Gengshan Yang, Deqing Sun, Varun Jampani, Daniel Vlasic, Forrester Cole, Huiwen Chang, Deva Ramanan, William T Freeman, and Ce Liu. Lasr: Learning articulated shape reconstruction from a monocular video. InCVPR, 2021

    2021

  49. [57]

    Banmo: Building animatable 3d neural models from many casual videos

    Gengshan Yang, Minh V o, Natalia Neverova, Deva Ramanan, Andrea Vedaldi, and Hanbyul Joo. Banmo: Building animatable 3d neural models from many casual videos. InCVPR, 2022

    2022

  50. [58]

    Lassie: Learning articulated shape from sparse image ensemble via 3d part discovery

    Chun-Han Yao, Wei-Chih Hung, Yuanzhen Li, Michael Rubinstein, Ming-Hsuan Yang, and Varun Jampani. Lassie: Learning articulated shape from sparse image ensemble via 3d part discovery. InNeurIPS, 2022

    2022

  51. [59]

    Deep part induction from articulated object pairs.ACM Transactions on Graphics (TOG), 37(6):1–15, 2018

    Li Yi, Haibin Huang, Difan Liu, Evangelos Kalogerakis, Hao Su, and Leonidas Guibas. Deep part induction from articulated object pairs.ACM Transactions on Graphics (TOG), 37(6):1–15, 2018

    2018

  52. [60]

    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. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 3947– 3956, 2019

    2019

  53. [61]

    Simple fast algorithms for the editing distance between trees and related problems.SIAM journal on computing, 18(6):1245–1262, 1989

    Kaizhong Zhang and Dennis Shasha. Simple fast algorithms for the editing distance between trees and related problems.SIAM journal on computing, 18(6):1245–1262, 1989

    1989

  54. [62]

    Open3D: A modern library for 3D data processing.arXiv:1801.09847, 2018

    Qian-Yi Zhou, Jaesik Park, and Vladlen Koltun. Open3D: A modern library for 3D data processing.arXiv:1801.09847, 2018. 16CHAO ET AL.: ARTICULA TED OBJECT PART SEGMENTA TION

    2018 arXiv

  55. [63]

    On the continuity of rotation representations in neural networks

    Yi Zhou, Connelly Barnes, Jingwan Lu, Jimei Yang, and Hao Li. On the continuity of rotation representations in neural networks. InProceedings of the IEEE/CVF confer- ence on computer vision and pattern recognition, pages 5745–5753, 2019

    2019

  56. [64]

    3d menagerie: Modeling the 3d shape and pose of animals

    Silvia Zuffi, Angjoo Kanazawa, David W Jacobs, and Michael J Black. 3d menagerie: Modeling the 3d shape and pose of animals. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 6365–6373, 2017

    2017

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.