Pith. sign in

REVIEW 5 major objections 7 minor 81 references

Level-Set Parameters: Novel Representation for 3D Shape Analysis

T0 review · 5 major / 7 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read SDF network weights can serve directly as a 3D shape representation, matching or exceeding point-cloud methods under arbitrary rotations and enabling correspondence-free 6D pose estimation.

desk verdict A genuinely new representation for 3D shapes with strong pose-robust results, but a few reproducibility gaps keep it from being fully convincing. read the letter →

arxiv 2412.13502 v2 pith:T4S2E4AA submitted 2024-12-18 cs.CV

Huan Lei , Hongdong Li , Andreas Geiger , Anthony Dick This is my paper · ORCID
classification cs.CV
keywords level-setparameterssigneddistancefunctions3Dshaperepresentationclassificationretrieval6Dobjectposeestimationhypernetworkneuralfields
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 that the learned weights of a signed distance function (SDF) network, called level-set parameters, can be used directly as a 3D shape representation instead of point clouds or meshes. Because these parameters are not Euclidean, the paper aligns shapes by decomposing them as $\theta = \mu + \Delta\theta$, where $\mu$ is a shared mean learned from a few shapes per category, treating the residuals as a pseudo-normal distribution. A hypernetwork conditions the first-layer parameters on rotation and translation, so the same representation supports arbitrary-pose classification, retrieval, and 6D pose estimation. The paper reports that this parameter-based representation matches or outperforms equivariant point-cloud networks under SO(3) rotations and yields accurate pose estimates from partial scans despite noise and outliers.

What carries the argument

The load-bearing identity is the pseudo-normal decomposition $\theta = \mu + \Delta\theta$, which reparameterizes SDF network weights as a shared mean $\mu$ plus per-shape residual $\Delta\theta$, emulating a normal distribution with identity covariance. The mean is learned in two stages: first a pose-conditioned hypernetwork (HyperSE3-SDF) trains $\mu$ and small latent matrices on a few shapes per category; then each shape's SDF is fit by training only $\Delta\theta$ from zero. The hypernetwork generates the first-layer weights and biases from $(R, t)$ through normalized linear combinations of latent matrices, satisfying the geometric initialization that SDF training needs and letting the same $\mu$ serve all poses. The encoder consumes three tensors formed from the layer parameters, and pose estimation optimizes $R$ and $t$ directly against the SDF reconstruction loss while the reference parameters stay frozen.

What would settle it

Retrain the two-stage construction with $\mu$ learned only from some categories and test classification on a held-out category; if accuracy collapses toward chance, the shared mean does not generalize across categories as claimed. A direct numerical check is to fit a multivariate normal to the empirical distribution of $\Delta\theta$ per category and measure its deviation from $\mathcal{N}(0, I)$: large deviations or category-dependent means would show the pseudo-normal assumption is not actually established by the data.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that a shape's identity and pose are recoverable from the weights of its individually overfit SDF network once those weights are put in a common coordinate frame. The authors construct this frame by learning a shared pose-dependent mean $\mu$ with a hypernetwork-augmented 8-layer SDF, then representing each shape by its residual $\Delta\theta = \theta - \mu$. Training an encoder on these residuals gives the highest accuracy among compared methods on Manifold40 under arbitrary rotations (87.02% OA versus 84.61% for a rotation-equivariant point-cloud baseline) and competitive results on ShapeNet. The same level-set parameters, kept frozen, let an optimizer recover a 6D pose from a partial point cloud by minimizing the SDF reconstruction loss, reporting errors below 1.4 degrees in rotation even with added noise and 30% outliers. The paper frames this as extending 3D analysis from discrete input data to a continuous parameter modality.

Load-bearing premise

The load-bearing premise is that a shared mean $\mu$ learned from a small number of shapes per category puts all shapes into a common parameter-space coordinate frame, so the residual parameters $\Delta\theta$, rather than the raw parameters $\theta$, carry the shape semantics; if that alignment fails, the classification and pose estimation built on $\Delta\theta$ would break.

Editorial extensions

If this is right

  • If level-set parameters carry shape semantics, classification and retrieval can be performed without point clouds or meshes, avoiding discrete-resolution artifacts and exhaustive rotation augmentation.
  • Because pose-dependent parameters are confined to the first SDF layer, a single shared mean supports arbitrary rotations, so training data can be augmented in parameter space rather than in 3D coordinate space.
  • 6D pose estimation reduces to optimizing rotation and translation against the SDF reconstruction loss, requiring no correspondences, global shape features, or training data for registration.
  • The same continuous representation should make shape analysis robust to input resolution and sampling density, since the surface is defined by a function's zero level set rather than by discrete samples.
  • The per-shape fitting cost drops substantially with the learned mean (about 4 minutes per shape versus 1 hour without it), making the construction of large level-set-parameter datasets practical.

Reading between the lines

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

  • Editorial inference: if $\theta = \mu + \Delta\theta$ truly aligns categories, the same alignment may power cross-category tasks such as few-shot classification or shape analogies, because residuals would live in a common parameter space.
  • Editorial inference: the pose-conditioning approach could extend to other neural field types, such as occupancy fields or radiance fields, whose parameters have known transformation behavior, though the paper only demonstrates SDFs.
  • Editorial inference: the paper's pose estimator assumes reference level-set parameters are already available; a testable extension is to jointly optimize a shape code and pose from a raw partial scan, removing that assumption.
  • Editorial inference: the method's reliance on a shared mean learned from 7 to 20 shapes per category could be stress-tested by learning the mean on one set of categories and evaluating on held-out categories, which the paper does not report.
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

5 major / 7 minor

Summary. The paper proposes using the weights of an 8-layer SDF MLP (called 'level-set parameters') as a standalone 3D shape representation. It decomposes these weights as θ = μ + Δθ, learns a shared pose-dependent mean μ with a small number of shapes via a hypernetwork, and treats the residual Δθ as the instance-level representation. An encoder trained on the three tensors derived from Δθ is used for classification and retrieval, and an SDF-reconstruction-loss optimization is used for 6D pose estimation from partial point clouds. Experiments on ShapeNet and Manifold40 report competitive or superior accuracy versus point-cloud baselines, notably 87.02% OA versus 84.61% for VN-DGCNN on Manifold40 under SO(3)/SO(3), and low rotation/translation errors on a 300-pair pose-estimation benchmark.

Significance. If sustained, the central proposal is significant: it extends 3D shape analysis to neural-network weight space and enables pose-conditioned reasoning without point-cloud augmentation or equivariant architectures. The strongest evidence is the external-baseline comparison in Table 3, the random-versus-learned μ ablation in Table 1, and the correspondence-free pose-estimation result in Table 4. The hypernetwork design that explicitly satisfies SAL initialization constraints is a principled construction. However, the lack of direct validation of the pseudo-normal residual model and the very small set of shapes used to train the shared mean μ leave the generalizability of the main mechanism under-supported.

major comments (5)
  1. [Section 3.2, Eq. (6)] The paper labels θ = μ + Δθ a 'pseudo-normal distribution' with identity covariance, but this is not compatible with the parameterization actually used: each component of Δθ is defined through tanh or a normalized inner product, so it is bounded in [-1,1] and cannot be a standard normal variable. Please report empirical diagnostics, including marginal histograms of Δθ, diagonal versus off-diagonal covariance energy, and per-category and per-pose variance, to show that the identity-covariance assumption is a reasonable approximation. This is not a philosophical objection: the normalization of θ by μ and the interpretation of Δθ as the instance-specific semantic content both depend on this assumption.
  2. [Section 3.2 and Section 5.1] The shared mean μ is trained on 20 shapes per class on ShapeNet and 7 per class on Manifold40, which is a very small sample from which to infer a category-wide, pose-generic alignment of a high-dimensional parameter space. The current validation is indirect: t-SNE in Fig. 3, Table 1, and downstream classification accuracy. Please add (a) a repeated-subset sensitivity analysis of μ, (b) the variance and geometry of Δθ on held-out shapes, and (c) results obtained when the μ-training shapes are excluded from the encoder training set. Without such evidence, the possibility that the gains over random μ reflect memorization of the few training shapes rather than a generalizable parameter-space alignment is not excluded.
  3. [Section 5.2, Tables 2-3] The empirical evaluation compares LSPData only against point-cloud networks. Since the paper positions level-set parameters as an alternative to other continuous neural-field representations, the classification and retrieval claims should also be benchmarked against at least one neural-field weight-space or modulation-vector baseline, such as the modulation vectors of Functa [65], HyperDiffusion [16], or DeepSDF latent codes [7], under the same train/test protocol. This comparison is needed to substantiate the claim that the proposed residual Δθ is a particularly effective continuous representation for shape semantics.
  4. [Section 4.2, Table 4] The pose-estimation experiment is a proof of concept rather than a systematic evaluation: it uses 10 shapes per category, 300 pairs, and only three classical optimization-based baselines. The setting also assumes that ground-truth reference level-set parameters θ are provided, which is a strong practical assumption. Please report per-category results, characterize failure cases, evaluate sensitivity to the Chamfer-filter threshold and SDF fit quality, and include at least one learning-based registration comparison. The current table is suggestive but too narrow to support the general conclusion in Section 6.
  5. [Section 1 and Section 5] A motivating advantage claimed for level-set parameters is robustness to variations in input resolution, but no experiment varies the number of sampled surface points or the resolution of the input point cloud. Adding a resolution sweep, or explicitly qualifying this claim in the text, would align the experiments with the paper's stated motivation.
minor comments (7)
  1. [Section 5.2] The heading 'Retrival' should be 'Retrieval'.
  2. [References, [17]] The text repeatedly refers to 'Ramirez et al. [17]', but the reference list entry [17] is attributed to De Luigi, Cardace, and Spezialetti; please correct the citation and the in-text attribution.
  3. [Section 3.2] 'This facilities the acquisition' should read 'This facilitates the acquisition'.
  4. [Section 5.2] The exact train/validation/test split for classification and retrieval, the number of encoder training epochs, and the selected values of λ1-λ4 and λreg are not stated; these details are needed for reproducibility.
  5. [Figure 3] Please state the number of points shown per category in the t-SNE plot and whether the parameters are visualized at a fixed pose or averaged over poses.
  6. [Table 5] Table 5 has a formatting problem: the columns under 'Method' are not readable as printed; please restructure the table.
  7. [Section 3.1] The statement that Eq. (4) yields 'shape semantics incomparable' to HyperSE3-SDF is vague; please specify the quantitative criterion, such as classification accuracy or reconstruction quality, used for the comparison.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the level-set parameter representation is benchmarked against external baselines and the pose results are obtained by test-time optimization, not by reusing fitted constants.

full rationale

Walking the derivation chain: per-shape SDF parameters θ are obtained by unsupervised reconstruction; the decomposition θ = μ + Δθ defines Δθ as the residual after a shared, unsupervised alignment mean μ. The encoder is trained on Δθ and evaluated on classification/retrieval against external point-cloud baselines (PointNet, DGCNN, VN-DGCNN) under held-out pose conditions, so the reported accuracies are not forced by construction. The ablation in Table 1 (random μ: 41.77 vs learned μ: 97.0) is an empirical comparison, not a tautology; a memorized μ that failed to generalize would show up as poor held-out accuracy. The pose-estimation section optimizes the SDF reconstruction loss at test time with the reference θ frozen, and is evaluated on separately generated partial point clouds with noise and outliers against ICP, FGR, and TEASER++; the small RRE/RTE values are optimization outcomes, not fitted predictions. The SAL initialization constraint (Eq. 15 and the text 'satisfy the geometric initializations recommended by SAL') is an external, parameter-free initialization rule, not an assertion of the paper's central claims. The only author self-citations ([4], [38]) appear in related-work enumerations and are not load-bearing. The stated limitation that level-set parameters are unsuitable for local features is a scope restriction, not a circular step. In short, no equation or fitted quantity is reused as its own prediction.

Assumptions & free parameters 8 free parameters · 5 assumptions · 0 invented entities

No new physical entities, forces, particles, or dimensions are introduced. The hypernetwork and LSPData are methods and datasets, not invented entities. The 'pseudo-normal distribution' is a modeling assumption and is captured under axioms.

free parameters (8)
  • Shared mean µ = learned from 20 shapes/class (ShapeNet), 7/class (Manifold40)
    The shared SDF-weight mean aligns all shapes in parameter space; its quality is central to the method and is learned from a small hand-picked number of shapes per category.
  • Loss weights λ1-λ4
    Balancing constants for Lp_dist, Ln_dist, Leik, Lp_norm in Eq. (2); no values are reported in the paper.
  • Regularization weight λreg
    L1 penalty in Eq. (11) encouraging ∆θ near zero; no value reported.
  • Latent matrix dimensions I, J = I=2, J=8
    Dimensions of the per-parameter latent matrices Ymn in the hypernetwork; chosen by hand.
  • Training epochs and batch size for stage one = 50000 epochs, batch 50
    Reported in Appendix C; hyperparameters chosen for convergence.
  • Pose search grid T, S, N, M = T=15, S=20, N=20, M=10
    Grid resolution, number of candidates, alternating rounds, and optimization steps for pose estimation.
  • Augmentation sigmas and dropout = σ1=0.2, σ2=0.05, dropout 0.5, σc=0.1
    Data augmentation for level-set parameter tensors in semantic analysis.
  • Chamfer distance filter threshold
    Shapes are filtered by unreported Chamfer distance thresholds to ensure surface quality (Section 5); this selection could affect benchmark results.
assumptions (5)
  • domain assumption An 8-layer MLP SDF network can represent each 3D shape accurately enough for classification and pose estimation.
    Invoked throughout Section 3; the whole parameter-space representation inherits the approximation quality of per-shape SDF fitting.
  • domain assumption A shared mean µ learned from a few shapes per category generalizes across all categories and poses.
    Section 3.2; the parameter-space alignment underlying ∆θ is validated only indirectly via t-SNE and classification accuracy.
  • standard math Rotations and translations can be absorbed by modifying only the first layer of the SDF MLP.
    Eq. (4) is exact for a perfect SDF, but the SDF is approximate; the paper assumes the approximation error is small enough that first-layer conditioning suffices (Section 3.1).
  • ad hoc to paper The decomposition θ = µ + ∆θ with identity covariance is a meaningful alignment of shapes.
    Section 3; the 'pseudo-normal' distribution is asserted, not measured; ∆θ is bounded by tanh, so it cannot be Gaussian.
  • domain assumption Chamfer-distance filtering removes only low-quality reconstructions and does not bias the semantic or pose comparisons.
    Section 5; filtering is applied to LSPData, and it is not stated whether baseline point-cloud methods see the same filtered split.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Level-Set Parameters: Novel Representation for 3D Shape Analysis." pith.science (2026). https://pith.science/paper/T4S2E4AA

@misc{pith2026241213502,
  author       = {Pith},
  title        = {Pith review of: Level-Set Parameters: Novel Representation for 3D Shape Analysis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/T4S2E4AA}},
  note         = {Machine review of arXiv:2412.13502}
}
read the original abstract

3D shape analysis has been largely focused on traditional 3D representations of point clouds and meshes, but the discrete nature of these data makes the analysis susceptible to variations in input resolutions. Recent development of neural fields brings in level-set parameters from signed distance functions as a novel, continuous, and numerical representation of 3D shapes, where the shape surfaces are defined as zero-level-sets of those functions. This motivates us to extend shape analysis from the traditional 3D data to these novel parameter data. Since the level-set parameters are not Euclidean like point clouds, we establish correlations across different shapes by formulating them as a pseudo-normal distribution, and learn the distribution prior from the respective dataset. To further explore the level-set parameters with shape transformations, we propose to condition a subset of these parameters on rotations and translations, and generate them with a hypernetwork. This simplifies the pose-related shape analysis compared to using traditional data. We demonstrate the promise of the novel representations through applications in shape classification (arbitrary poses), retrieval, and 6D object pose estimation.

Figures

Figures reproduced from arXiv: 2412.13502 by the authors.

Figure 1
Figure 1. Hypernetwork for surface transformation. The hypernetwork [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Encoder-based semantic learning from level-set parameters. The network in (a) processes [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. t-SNE embeddings of level-set parameter data obtained with different [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Surface quality of ShapeNet. We apply the level-set parameters to semantic analysis on ShapeNet and Manifold40. For time concern, a maximum of 2000 shapes are reconstructed for each class in ShapeNet. We filter the continuous shapes based on their Chamfer dis￾tances to…
Figure 5
Figure 5. Figure 5: Registration of Teaser++ and ours. RRE/RTE metrics are shown, with RTE scaled by [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: The SDF network and its resulting level-set parameters for shape representation. [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]
Figure 7
Figure 7. Figure 7: Comparison of different hypernetworks at transforming the continuous shape surfaces. [PITH_FULL_IMAGE:figures/full_fig_p018_7.png]
Figure 8
Figure 8. Figure 8: Distributions of Euler Angles for the ground-truth rotations. [PITH_FULL_IMAGE:figures/full_fig_p019_8.png]
Figure 9
Figure 9. Figure 9: Registration comparison between Teaser++ and the proposed method. We report RRE/RTE [PITH_FULL_IMAGE:figures/full_fig_p020_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

81 extracted references · 58 canonical work pages

  1. [65]

    From data to functa: Your data point is a function and you can treat it like one

    Emilien Dupont, Hyunjik Kim, SM Eslami, Danilo Rezende, and Dan Rosenbaum. From data to functa: Your data point is a function and you can treat it like one. arXiv preprint arXiv:2201.12204, 2022

    arXiv 2022

  2. [16]

    Hyperdiffusion: Generating im- plicit neural fields with weight-space diffusion

    Ziya Erkoç, Fangchang Ma, Qi Shan, Matthias Nießner, and Angela Dai. Hyperdiffusion: Generating im- plicit neural fields with weight-space diffusion. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 14300–14310, 2023

    work page 2023

  3. [7]

    Deepsdf: Learning continuous signed distance functions for shape representation

    Jeong Joon Park, Peter Florence, Julian Straub, Richard Newcombe, and Steven Lovegrove. Deepsdf: Learning continuous signed distance functions for shape representation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 165–174, 2019

    2019

  4. [1]

    Pointnet: Deep learning on point sets for 3d classification and segmentation

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

    2017

  5. [2]

    Dynamic graph cnn for learning on point clouds

    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 Transactions on Graphics (tog), 38(5):1–12, 2019

    2019

  6. [3]

    PointCNN: Convolution on x-transformed points

    Yangyan Li, Rui Bu, Mingchao Sun, Wei Wu, Xinhan Di, and Baoquan Chen. PointCNN: Convolution on x-transformed points. In Advances in Neural Information Processing Systems, pages 820–830, 2018

    work page 2018

  7. [4]

    Spherical kernel for efficient graph convolution on 3d point clouds

    Huan Lei, Naveed Akhtar, and Ajmal Mian. Spherical kernel for efficient graph convolution on 3d point clouds. IEEE transactions on pattern analysis and machine intelligence, 43(10):3664–3680, 2020

    work page 2020

  8. [5]

    MeshCNN: a network with an edge

    Rana Hanocka, Amir Hertz, Noa Fish, Raja Giryes, Shachar Fleishman, and Daniel Cohen-Or. MeshCNN: a network with an edge. ACM Transactions on Graphics (ToG), 38(4):1–12, 2019

    work page 2019

Show all 81 references
  1. [6]

    Subdivision-based mesh convolution networks

    Shi-Min Hu, Zheng-Ning Liu, Meng-Hao Guo, Jun-Xiong Cai, Jiahui Huang, Tai-Jiang Mu, and Ralph R Martin. Subdivision-based mesh convolution networks. ACM Transactions on Graphics (TOG), 41(3): 1–16, 2022

    2022

  2. [8]

    Implicit neural representations with periodic activation functions

    Vincent Sitzmann, Julien Martel, Alexander Bergman, David Lindell, and Gordon Wetzstein. Implicit neural representations with periodic activation functions. Advances in neural information processing systems, 33:7462–7473, 2020

    2020

  3. [9]

    Neural fields in visual computing and beyond

    Yiheng Xie, Towaki Takikawa, Shunsuke Saito, Or Litany, Shiqin Yan, Numair Khan, Federico Tombari, James Tompkin, Vincent Sitzmann, and Srinath Sridhar. Neural fields in visual computing and beyond. In Computer Graphics Forum, volume 41, pages 641–676. Wiley Online Library, 2022

    2022

  4. [10]

    Occupancy networks: Learning 3d reconstruction in function space

    Lars Mescheder, Michael Oechsle, Michael Niemeyer, Sebastian Nowozin, and Andreas Geiger. Occupancy networks: Learning 3d reconstruction in function space. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4460–4470, 2019

    2019

  5. [11]

    Points2surf learning implicit surfaces from point clouds

    Philipp Erler, Paul Guerrero, Stefan Ohrhallinger, Niloy J Mitra, and Michael Wimmer. Points2surf learning implicit surfaces from point clouds. In European Conference on Computer Vision, pages 108–124. Springer, 2020

    2020

  6. [12]

    Learning implicit fields for generative shape modeling

    Zhiqin Chen and Hao Zhang. Learning implicit fields for generative shape modeling. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5939–5948, 2019

    2019

  7. [13]

    Convolutional occupancy networks

    Songyou Peng, Michael Niemeyer, Lars Mescheder, Marc Pollefeys, and Andreas Geiger. Convolutional occupancy networks. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part III 16, pages 523–540. Springer, 2020

    2020

  8. [14]

    Implicit functions in feature space for 3d shape reconstruction and completion

    Julian Chibane, Thiemo Alldieck, and Gerard Pons-Moll. Implicit functions in feature space for 3d shape reconstruction and completion. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 6970–6981, 2020

    2020

  9. [15]

    MetaSDF: Meta- learning signed distance functions

    Vincent Sitzmann, Eric Chan, Richard Tucker, Noah Snavely, and Gordon Wetzstein. MetaSDF: Meta- learning signed distance functions. Advances in Neural Information Processing Systems, 33:10136–10147, 2020

    2020

  10. [17]

    Deep learning on 3D neural fields

    Luca De Luigi, Adriano Cardace, and Riccardo Spezialetti. Deep learning on 3D neural fields. 2023

    2023

  11. [18]

    Vector neurons: A general framework for so (3)-equivariant networks

    Congyue Deng, Or Litany, Yueqi Duan, Adrien Poulenard, Andrea Tagliasacchi, and Leonidas J Guibas. Vector neurons: A general framework for so (3)-equivariant networks. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 12200–12209, 2021

    2021

  12. [19]

    Learning so (3) equivariant representations with spherical cnns

    Carlos Esteves, Christine Allen-Blanchette, Ameesh Makadia, and Kostas Daniilidis. Learning so (3) equivariant representations with spherical cnns. In Proceedings of the European Conference on Computer Vision (ECCV), pages 52–68, 2018. 10

    2018

  13. [20]

    Spherical cnns

    Taco S Cohen, Mario Geiger, Jonas Köhler, and Max Welling. Spherical cnns. arXiv preprint arXiv:1801.10130, 2018

    2018 arXiv

  14. [21]

    Tensor field networks: Rotation-and translation-equivariant neural networks for 3D point clouds

    Nathaniel Thomas, Tess Smidt, Steven Kearnes, Lusann Yang, Li Li, Kai Kohlhoff, and Patrick Riley. Tensor field networks: Rotation-and translation-equivariant neural networks for 3D point clouds. arXiv preprint arXiv:1802.08219, 2018

    2018 arXiv

  15. [22]

    A functional approach to rotation equivariant non-linearities for tensor field networks

    Adrien Poulenard and Leonidas J Guibas. A functional approach to rotation equivariant non-linearities for tensor field networks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13174–13183, 2021

    2021

  16. [23]

    Shapenet: An information-rich 3d model repository

    Angel X Chang, Thomas Funkhouser, Leonidas Guibas, Pat Hanrahan, Qixing Huang, Zimo Li, Silvio Savarese, Manolis Savva, Shuran Song, Hao Su, et al. Shapenet: An information-rich 3d model repository. arXiv preprint arXiv:1512.03012, 2015

    2015 arXiv

  17. [24]

    3d shapenets: A deep representation for volumetric shapes

    Zhirong Wu, Shuran Song, Aditya Khosla, Fisher Yu, Linguang Zhang, Xiaoou Tang, and Jianxiong Xiao. 3d shapenets: A deep representation for volumetric shapes. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1912–1920, 2015

    1912

  18. [25]

    Learning-based point cloud registration for 6d object pose estimation in the real world

    Zheng Dang, Lizhou Wang, Yu Guo, and Mathieu Salzmann. Learning-based point cloud registration for 6d object pose estimation in the real world. In European conference on computer vision, pages 19–37. Springer, 2022

    2022

  19. [26]

    Escape from cells: Deep kd-networks for the recognition of 3d point cloud models

    Roman Klokov and Victor Lempitsky. Escape from cells: Deep kd-networks for the recognition of 3d point cloud models. In Proceedings of the IEEE International Conference on Computer Vision, pages 863–872. IEEE, 2017

    2017

  20. [27]

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

    Charles R Qi, Li Yi, Hao Su, and Leonidas J Guibas. PointNet++: Deep hierarchical feature learning on point sets in a metric space. Advances in Neural Information Processing Systems, 2017

    2017

  21. [28]

    Graph attention convolution for point cloud semantic segmentation

    Lei Wang, Yuchun Huang, Yaolin Hou, Shenman Zhang, and Jie Shan. Graph attention convolution for point cloud semantic segmentation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10296–10305, 2019

    2019

  22. [29]

    Pointconv: Deep convolutional networks on 3d point clouds

    Wenxuan Wu, Zhongang Qi, and Li Fuxin. Pointconv: Deep convolutional networks on 3d point clouds. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 9621–9630, 2019

    2019

  23. [30]

    Qi, Jean-Emmanuel Deschaud, Beatriz Marcotegui, François Goulette, and Leonidas J

    Hugues Thomas, Charles R. Qi, Jean-Emmanuel Deschaud, Beatriz Marcotegui, François Goulette, and Leonidas J. Guibas. Kpconv: Flexible and deformable convolution for point clouds. Proceedings of the IEEE International Conference on Computer Vision, 2019

    2019

  24. [31]

    Point transformer

    Hengshuang Zhao, Li Jiang, Jiaya Jia, Philip HS Torr, and Vladlen Koltun. Point transformer. In Proceedings of the IEEE/CVF international conference on computer vision, pages 16259–16268, 2021

    2021

  25. [32]

    PCT: Point cloud transformer

    Meng-Hao Guo, Jun-Xiong Cai, Zheng-Ning Liu, Tai-Jiang Mu, Ralph R Martin, and Shi-Min Hu. PCT: Point cloud transformer. Computational Visual Media, 7:187–199, 2021

    2021

  26. [33]

    Point transformer V2: Grouped vector attention and partition-based pooling

    Xiaoyang Wu, Yixing Lao, Li Jiang, Xihui Liu, and Hengshuang Zhao. Point transformer V2: Grouped vector attention and partition-based pooling. Advances in Neural Information Processing Systems, 35: 33330–33342, 2022

    2022

  27. [34]

    V oxnet: A 3d convolutional neural network for real-time object recognition

    Daniel Maturana and Sebastian Scherer. V oxnet: A 3d convolutional neural network for real-time object recognition. In 2015 IEEE/RSJ international conference on intelligent robots and systems (IROS), pages 922–928. IEEE, 2015

    2015

  28. [35]

    Octnet: Learning deep 3d representations at high resolutions

    Gernot Riegler, Ali Osman Ulusoy, and Andreas Geiger. Octnet: Learning deep 3d representations at high resolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3577–3586, 2017

    2017

  29. [36]

    3d semantic segmentation with submanifold sparse convolutional networks

    Benjamin Graham, Martin Engelcke, and Laurens van der Maaten. 3d semantic segmentation with submanifold sparse convolutional networks. CVPR, 2018

    2018

  30. [37]

    Hodgenet: Learning spectral geometry on triangle meshes

    Dmitriy Smirnov and Justin Solomon. Hodgenet: Learning spectral geometry on triangle meshes. ACM Transactions on Graphics (TOG), 40(4):1–11, 2021

    2021

  31. [38]

    Mesh convolution with continuous filters for 3-d surface parsing

    Huan Lei, Naveed Akhtar, Mubarak Shah, and Ajmal Mian. Mesh convolution with continuous filters for 3-d surface parsing. IEEE Transactions on Neural Networks and Learning Systems, 2023. 11

    2023

  32. [39]

    Deep object pose estimation for semantic robotic grasping of household objects.arXiv preprint arXiv:1809.10790, 2018

    Jonathan Tremblay, Thang To, Balakumar Sundaralingam, Yu Xiang, Dieter Fox, and Stan Birchfield. Deep object pose estimation for semantic robotic grasping of household objects.arXiv preprint arXiv:1809.10790, 2018

    2018 arXiv

  33. [40]

    Pose estimation for augmented reality: a hands-on survey

    Eric Marchand, Hideaki Uchiyama, and Fabien Spindler. Pose estimation for augmented reality: a hands-on survey. IEEE transactions on visualization and computer graphics, 22(12):2633–2651, 2015

    2015

  34. [41]

    Are we ready for autonomous driving? the kitti vision benchmark suite

    Andreas Geiger, Philip Lenz, and Raquel Urtasun. Are we ready for autonomous driving? the kitti vision benchmark suite. In 2012 IEEE conference on computer vision and pattern recognition, pages 3354–3361. IEEE, 2012

    2012

  35. [42]

    Densefusion: 6d object pose estimation by iterative dense fusion

    Chen Wang, Danfei Xu, Yuke Zhu, Roberto Martín-Martín, Cewu Lu, Li Fei-Fei, and Silvio Savarese. Densefusion: 6d object pose estimation by iterative dense fusion. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 3343–3352, 2019

    2019

  36. [43]

    Normal- ized object coordinate space for category-level 6d object pose and size estimation

    He Wang, Srinath Sridhar, Jingwei Huang, Julien Valentin, Shuran Song, and Leonidas J Guibas. Normal- ized object coordinate space for category-level 6d object pose and size estimation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages...

    2019

  37. [44]

    PVnet: Pixel-wise voting network for 6dof pose estimation

    Sida Peng, Yuan Liu, Qixing Huang, Xiaowei Zhou, and Hujun Bao. PVnet: Pixel-wise voting network for 6dof pose estimation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4561–4570, 2019

    2019

  38. [45]

    Pix2pose: Pixel-wise coordinate regression of objects for 6d pose estimation

    Kiru Park, Timothy Patten, and Markus Vincze. Pix2pose: Pixel-wise coordinate regression of objects for 6d pose estimation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 7668–7677, 2019

    2019

  39. [46]

    Center-based decoupled point-cloud registration for 6d object pose estimation

    Haobo Jiang, Zheng Dang, Shuo Gu, Jin Xie, Mathieu Salzmann, and Jian Yang. Center-based decoupled point-cloud registration for 6d object pose estimation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 3427–3437, 2023

    2023

  40. [47]

    3dmatch: Learning local geometric descriptors from rgb-d reconstructions

    Andy Zeng, Shuran Song, Matthias Nießner, Matthew Fisher, Jianxiong Xiao, and Thomas Funkhouser. 3dmatch: Learning local geometric descriptors from rgb-d reconstructions. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1802–1811, 2017

    2017

  41. [48]

    Fully convolutional geometric features

    Christopher Choy, Jaesik Park, and Vladlen Koltun. Fully convolutional geometric features. InProceedings of the IEEE/CVF international conference on computer vision, pages 8958–8966, 2019

    2019

  42. [49]

    Deep closest point: Learning representations for point cloud registration

    Yue Wang and Justin M Solomon. Deep closest point: Learning representations for point cloud registration. In Proceedings of the IEEE/CVF international conference on computer vision, pages 3523–3532, 2019

    2019

  43. [50]

    Predator: Registration of 3d point clouds with low overlap

    Shengyu Huang, Zan Gojcic, Mikhail Usvyatsov, Andreas Wieser, and Konrad Schindler. Predator: Registration of 3d point clouds with low overlap. InProceedings of the IEEE/CVF Conference on computer vision and pattern recognition, pages 4267–4276, 2021

    2021

  44. [51]

    Buffer: Balancing accuracy, efficiency, and generalizability in point cloud registration

    Sheng Ao, Qingyong Hu, Hanyun Wang, Kai Xu, and Yulan Guo. Buffer: Balancing accuracy, efficiency, and generalizability in point cloud registration. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 1255–1264, 2023

    2023

  45. [52]

    Feature-metric registration: A fast semi-supervised approach for robust point cloud registration without correspondences

    Xiaoshui Huang, Guofeng Mei, and Jian Zhang. Feature-metric registration: A fast semi-supervised approach for robust point cloud registration without correspondences. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 11366–11374, 2020

    2020

  46. [53]

    Pointnetlk: Robust & efficient point cloud registration using pointnet

    Yasuhiro Aoki, Hunter Goforth, Rangaprasad Arun Srivatsan, and Simon Lucey. Pointnetlk: Robust & efficient point cloud registration using pointnet. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 7163–7172, 2019

    2019

  47. [54]

    Method for registration of 3-d shapes

    Paul J Besl and Neil D McKay. Method for registration of 3-d shapes. In Sensor fusion IV: control paradigms and data structures, volume 1611, pages 586–606. Spie, 1992

    1992

  48. [55]

    Fast global registration

    Qian-Yi Zhou, Jaesik Park, and Vladlen Koltun. Fast global registration. In Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part II 14, pages 766–782. Springer, 2016

    2016

  49. [56]

    Teaser: Fast and certifiable point cloud registration

    Heng Yang, Jingnan Shi, and Luca Carlone. Teaser: Fast and certifiable point cloud registration. IEEE Transactions on Robotics, 37(2):314–333, 2020

    2020

  50. [57]

    Go-icp: A globally optimal solution to 3d icp point-set registration

    Jiaolong Yang, Hongdong Li, Dylan Campbell, and Yunde Jia. Go-icp: A globally optimal solution to 3d icp point-set registration. IEEE transactions on pattern analysis and machine intelligence , 38(11): 2241–2254, 2015. 12

    2015

  51. [58]

    SAL: Sign agnostic learning of shapes from raw data

    Matan Atzmon and Yaron Lipman. SAL: Sign agnostic learning of shapes from raw data. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 2565–2574, 2020

    2020

  52. [59]

    Implicit geometric regularization for learning shapes

    Amos Gropp, Lior Yariv, Niv Haim, Matan Atzmon, and Yaron Lipman. Implicit geometric regularization for learning shapes. In International Conference on Machine Learning, pages 3789–3799. PMLR, 2020

    2020

  53. [60]

    Nerf in the wild: Neural radiance fields for unconstrained photo collections

    Ricardo Martin-Brualla, Noha Radwan, Mehdi SM Sajjadi, Jonathan T Barron, Alexey Dosovitskiy, and Daniel Duckworth. Nerf in the wild: Neural radiance fields for unconstrained photo collections. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognitio...

    2021

  54. [61]

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

    Peng Wang, Lingjie Liu, Yuan Liu, Christian Theobalt, Taku Komura, and Wenping Wang. Neus: Learning neural implicit surfaces by volume rendering for multi-view reconstruction. 34:27171–27183, 2021

    2021

  55. [62]

    MonoSDF: Exploring monocular geometric cues for neural implicit surface reconstruction

    Zehao Yu, Songyou Peng, Michael Niemeyer, Torsten Sattler, and Andreas Geiger. MonoSDF: Exploring monocular geometric cues for neural implicit surface reconstruction. Advances in neural information processing systems, 35:25018–25032, 2022

    2022

  56. [63]

    V olume rendering of neural implicit surfaces

    Lior Yariv, Jiatao Gu, Yoni Kasten, and Yaron Lipman. V olume rendering of neural implicit surfaces. Advances in Neural Information Processing Systems, 34:4805–4815, 2021

    2021

  57. [64]

    Digs: Divergence guided shape implicit neural representation for unoriented point clouds

    Yizhak Ben-Shabat, Chamin Hewa Koneputugodage, and Stephen Gould. Digs: Divergence guided shape implicit neural representation for unoriented point clouds. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 19323–19332, 2022

    2022

  58. [66]

    Modulated periodic activations for generalizable local functional representations

    Ishit Mehta, Michaël Gharbi, Connelly Barnes, Eli Shechtman, Ravi Ramamoorthi, and Manmohan Chan- draker. Modulated periodic activations for generalizable local functional representations. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 14214–...

    2021

  59. [67]

    pi-gan: Periodic implicit generative adversarial networks for 3d-aware image synthesis

    Eric R Chan, Marco Monteiro, Petr Kellnhofer, Jiajun Wu, and Gordon Wetzstein. pi-gan: Periodic implicit generative adversarial networks for 3d-aware image synthesis. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5799–5809, 2021

    2021

  60. [68]

    Model-agnostic meta-learning for fast adaptation of deep networks

    Chelsea Finn, Pieter Abbeel, and Sergey Levine. Model-agnostic meta-learning for fast adaptation of deep networks. In International conference on machine learning, pages 1126–1135. PMLR, 2017

    2017

  61. [69]

    Learned initializations for optimizing coordinate-based neural representations

    Matthew Tancik, Ben Mildenhall, Terrance Wang, Divi Schmidt, Pratul P Srinivasan, Jonathan T Barron, and Ren Ng. Learned initializations for optimizing coordinate-based neural representations. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition...

    2021

  62. [70]

    An introduction to variational autoencoders

    Diederik P Kingma, Max Welling, et al. An introduction to variational autoencoders. Foundations and Trends® in Machine Learning, 12(4):307–392, 2019

    2019

  63. [71]

    Gensdf: Two-stage learning of generalizable signed distance functions

    Gene Chou, Ilya Chugunov, and Felix Heide. Gensdf: Two-stage learning of generalizable signed distance functions. Advances in Neural Information Processing Systems, 35:24905–24919, 2022

    2022

  64. [72]

    Batch normalization: Accelerating deep network training by reducing internal covariate shift

    Sergey Ioffe and Christian Szegedy. Batch normalization: Accelerating deep network training by reducing internal covariate shift. In International conference on machine learning, pages 448–456. pmlr, 2015

    2015

  65. [73]

    Srinivasan, Matthew Tancik, Jonathan T

    Ben Mildenhall, Pratul P. Srinivasan, Matthew Tancik, Jonathan T. Barron, Ravi Ramamoorthi, and Ren Ng. NeRF: Representing scenes as neural radiance fields for view synthesis. In European conference on computer vision, pages 405–421, 2020

    2020

  66. [74]

    Visualizing data using t-sne

    Laurens Van der Maaten and Geoffrey Hinton. Visualizing data using t-sne. Journal of machine learning research, 9(11), 2008

    2008

  67. [75]

    Direct visibility of point sets

    Sagi Katz, Ayellet Tal, and Ronen Basri. Direct visibility of point sets. In ACM SIGGRAPH 2007 papers, pages 24–es. 2007

    2007

  68. [76]

    Spatial functa: Scaling functa to imagenet classification and generation

    Matthias Bauer, Emilien Dupont, Andy Brock, Dan Rosenbaum, Jonathan Richard Schwarz, and Hyun- jik Kim. Spatial functa: Scaling functa to imagenet classification and generation. arXiv preprint arXiv:2302.03130, 2023

    2023 arXiv

  69. [77]

    Generative neural fields by mixtures of neural implicit functions

    Tackgeun You, Mijeong Kim, Jungtaek Kim, and Bohyung Han. Generative neural fields by mixtures of neural implicit functions. Advances in Neural Information Processing Systems, 36, 2024. 13

    2024

  70. [78]

    Machine learning: a probabilistic perspective

    Kevin P Murphy. Machine learning: a probabilistic perspective. MIT press, 2012

    2012

  71. [79]

    Deep neural networks as gaussian processes

    Jaehoon Lee, Yasaman Bahri, Roman Novak, Samuel S Schoenholz, Jeffrey Pennington, and Jascha Sohl-Dickstein. Deep neural networks as gaussian processes. international conference on learning representations, 2017

    2017

  72. [80]

    Adam: A method for stochastic optimization

    Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014

    2014 arXiv

  73. [81]

    Rotation and translation invariant representation learning with implicit neural representations

    Sehyun Kwon, Joo Young Choi, and Ernest K Ryu. Rotation and translation invariant representation learning with implicit neural representations. In International Conference on Machine Learning, pages 18037–18056. PMLR, 2023. 14 A The SDF Network A.1 Tensors of Level-set Paramet...

    2023

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.