Pith. sign in

REVIEW 3 major objections 4 minor 43 references

FA-KPConv: Introducing Euclidean Symmetries to KPConv via Frame Averaging

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

Pith's one-line read Wrapping KPConv-based point-cloud networks in frame averaging yields exact invariance and equivariance to Euclidean transformations without adding any learnable parameters.

desk verdict Sound FA-on-KPConv engineering, but the 'simply wrapping' slogan oversells it; the joint-action invariance is correct, the feature-handling story is not. read the letter →

arxiv 2505.04485 v2 pith:TJK7QH5W submitted 2025-05-07 cs.CV

classification cs.CV
keywords FrameAveragingKPConvpointcloudclassificationregistrationequivarianceinvarianceEuclideangroup3Ddeeplearning
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 claims that any point-cloud network built from KPConv layers can be made exactly invariant or equivariant to translations, rotations, and reflections by wrapping it in Frame Averaging, without adding a single learnable parameter. The construction replaces the intractable average over the whole Euclidean group with an average over a small input-dependent frame: the orthogonal frame obtained from the eigendecomposition of the point cloud's covariance matrix. The authors argue that this embeds geometric prior knowledge directly into the network, so classification and registration improve precisely in the regimes where such priors matter, namely rotated test data and scarce training data, while performance on already-aligned data stays roughly unchanged. The paper demonstrates this on ModelNet40 classification and 3DMatch/3DLoMatch registration by comparing KPConv baselines with their FA-KPConv counterparts.

What carries the argument

The load-bearing object is the frame F(X): for a point cloud X, one computes the centroid c and covariance matrix C, then takes the eigendecomposition C = Q\Lambda Q^T, whose unit eigenvectors define an orthogonal frame up to sign choices. The set of all sign choices, together with the centering translation, forms a frame for the desired group, with size 1 for translations, 4 for rotations, 8 for rotations and reflections, and similarly for their combinations with translations. Frame averaging then replaces the intractable group average with an average over this small set, keeping the symmetrization exact. The paper wraps KPConv networks with this averaging, requiring the feature dimension to be a multiple of 3 so that $g^{{-1}}$ can act on Fin, and using a constant vector 1 in $R^{3}$ as the baseline input feature.

What would settle it

Take a KPConv network with non-geometric input features, apply a random rotation to both the coordinates and the features, and check whether the frame-averaged output is exactly unchanged; any rotation that changes the output would falsify the claimed exact invariance for the general f(X, Fin) case. A second test is to use point clouds with degenerate covariance spectra, such as planar or collinear points, where the eigendecomposition frame is not unique and the averaging may become fragile.

Watch

Extended reading notes

Core claim

The central claim is that Frame Averaging can be applied directly to a KPConv network f(X, Fin), where X are point coordinates and Fin are input features, to produce a function that is exactly invariant or equivariant under simultaneous Euclidean transformations of X and Fin. Specifically, the paper defines the equivariant function \hat f and the invariant function \bar f by averaging f over the frame F(X) of the input cloud, as in equations (14) and (15). It claims that this symmetrization remains exact even when the network depends on both coordinates and features, not only on coordinates, and that this holds for any KPConv-based architecture, from a single convolution layer to full classification or registration networks. The cost is not in parameters, which stay unchanged, but in compute and memory, which grow by a factor of |F(X)|, up to 8 for the full Euclidean group in 3D.

Load-bearing premise

The load-bearing premise is that frame averaging remains exactly invariant or equivariant when the network consumes both coordinates X and input features Fin, a property the paper asserts was 'verified via extensive testing' without providing a proof or test details; if this premise fails, the advertised exact symmetry does not hold for networks that use features beyond coordinates.

Editorial extensions

If this is right

  • FA-KPConv models trained on a small fraction of ModelNet40 remain almost unaffected by random test-time rotation: with 9843 training samples, accuracy on rotated data is 87.0% versus 87.1% on the original data, while the baseline drops from 90.4% to 44.6%.
  • On the low-overlap 3DLoMatch benchmark with 1k training samples, FA-GeoTransformer improves inlier ratio by up to 19.7% over the baseline and reduces relative rotation error by about 9.9% on rotated test data.
  • Because no parameters are added, any performance gain must come from the embedded geometric prior rather than from increased model capacity.
  • In already-aligned, canonically oriented datasets the enforced invariance can slightly hurt, which the paper attributes to the model spending capacity to work around an unnecessary constraint.

Reading between the lines

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

  • A formal proof that the frame-averaged function remains exact for f(X, Fin) would require verifying that the frame is equivariant, F(gX) = gF(X), and that the feature action is linear; the paper's 'extensive testing' leaves this unproven but testable.
  • If exactness holds for arbitrary features, the same wrapping could be applied to other coordinate-based 3D backbones beyond KPConv, yielding exactly equivariant versions of sparse convolutions or graph networks at the same parameter count.
  • The paper's baseline replaces scalar input features with a constant 3-vector; a natural extension would test what happens when Fin contains non-geometric information such as color or semantic labels, where the symmetry action on features is not physically meaningful.
  • The |F(X)| compute multiplier could be reduced by selecting one canonical eigenvector orientation instead of averaging over all sign choices, at the price of exactness when eigenvalues are degenerate.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper proposes FA-KPConv, a wrapper around KPConv-based point-cloud networks that uses Frame Averaging (Puny et al.) to make the network exactly invariant or equivariant to translations, rotations, and reflections. The construction represents a network as Y = f(X, Fin), computes a frame from the covariance/eigendecomposition of the coordinates X, and replaces f by an average over the frame, after reshaping input features into d-dimensional blocks so that they transform under the same Euclidean action. Experiments on ModelNet40 classification and 3DMatch/3DLoMatch registration compare modified KPConv baselines using constant 3-vector features against their FA versions, reporting gains mainly on rotated test data and in the low-training-data regime.

Significance. If properly scoped, the contribution is useful: it provides a simple recipe for making KPConv-based networks exactly invariant under Euclidean motions for geometric vector features, with no additional trainable parameters, and the experiments support gains in the low-data and rotated-test regimes. The mathematical claim is correct under the joint action on X and Fin, contrary to one of the stress-test concerns, but the paper states this without proof and overstates the 'simple wrapping' and 'no compromise' aspects for non-geometric features. The rotated-test protocol is also ambiguous about whether features are transformed. These issues are fixable and do not invalidate the core construction, but they do affect the exactness claim as advertised.

major comments (3)
  1. [III-C (Eqs. 14-17)] The paper asserts that exact invariance holds when f depends on both X and Fin, but justifies this only with 'We verified this via extensive testing' in Section III-C. This is load-bearing because exactness is the central claim. Provide a short proof: for an equivariant frame satisfying F(hX) = hF(X), re-indexing g = h g' in the sum over F(hX) gives the claimed equality in Eq. (15); the equivariant case Eq. (14) follows similarly. The proof should also state explicitly that Fin is acted on by the same g; without that joint action, Eqs. (16)-(17) do not follow.
  2. [IV-A, IV-B and Abstract] The abstract's claim of 'simply wrapping around an existing KPConv-based network' is not accurate for the networks used in the experiments. The original KP-CNN and GeoTransformer use a constant scalar feature equal to 1 (cin = 1), which is invariant under rotations; the paper changes the input to a constant 3-vector and compares FA models against this modified baseline, not against the original network. Thus the experiments do not show that the original KPConv network can be wrapped without changing its input representation, and the claim 'not compromising any input information' is only true in the weak sense of preserving information content, not in the sense of preserving the original feature space. Please state the scope explicitly: exact invariance applies when Fin consists of geometric d-vectors that transform under the same Euclidean action.
  3. [IV (Experimental protocol)] The rotated-test protocol is ambiguous: Section IV says the test data are 'once on a rotated version of it, in which each sample is randomly rotated' but does not state whether the input features Fin (the vector-ones features) are rotated together with the coordinates X. Since Eq. (17) guarantees invariance only for the joint action (gX, gFin), the exact-invariance claim is not meaningfully evaluated unless the same g is applied to Fin in the rotated tests. Specify the exact transformation applied to X and Fin in the rotated experiments, and for registration state whether the source and target point clouds receive the same or independent random rotations.
minor comments (4)
  1. [III-C] Please clarify the reshaping used to apply g to Fin: the sentence requiring cin to be a multiple of d should explicitly define the reshape/respace operation and its inverse, since Eq. (14) uses the notation g·Fin without a formal definition.
  2. [Table II] Consider adding the results of the original unmodified KP-CNN with scalar constant features (cin = 1) to Table II, so readers can quantify the effect of the vector-one reparameterization separately from the effect of frame averaging.
  3. [V (Analysis)] The conjecture that FA models 'spend capacity on learning to ignore the invariances imposed by design' is speculative; it would be helpful to cite or propose a concrete diagnostic, such as probing the frame-averaged outputs on aligned versus rotated inputs.
  4. [Figure 1] The two diagrams in Figure 1 are nearly identical and the labels are small; enlarging the figure and highlighting the difference between the equivariant and invariant cases would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: frame averaging is applied as an external, parameter-free construction, and the claimed invariance is definitional rather than derived from fitted inputs.

full rationale

The claimed derivation is the standard Frame Averaging construction of Eqs. (14)-(15). No quantity is fitted to the benchmarks and no parameter is renamed as a prediction. The invariance of Eq. (15) under the joint action (X,Fin)->(gX,gFin) follows from the equivariance of the frame F(X) established in [8] together with the group-averaging identity; the paper's phrase 'We verified this via extensive testing' is a missing proof, not a circular step, because the result does not assume the conclusion. The self-citation [41] (Rath and Condurache, with Condurache as a co-author) is used only for textbook definitions of equivariance and group averaging and is not load-bearing. The benchmarks (ModelNet40, 3DMatch) and the base architectures (KPConv, GeoTransformer) are external. The main caveat is that the experiments alter the baseline input feature from a scalar 1 to a vector 1 in R3 to satisfy cin=kd; this means the advertised 'simply wrapping ... not compromising any input information' is an overstatement for the original scalar-feature network, but this is a scope/correctness issue rather than circularity. Therefore no circular step is found.

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

Free parameters: none; the method introduces no learnable parameters. Axioms: the FA theorem from prior work, the unproved joint X-and-Fin extension, and the non-degenerate covariance assumption. Invented entities: none.

assumptions (3)
  • standard math Frame averaging over the covariance-based frame yields exact averaging over the full Euclidean group (the FA theorem from Puny et al.).
    Inherited from the cited FA paper; invoked in Section III-A when Equations (5)-(6) are replaced by (14)-(15).
  • ad hoc to paper Frame averaging remains exact when the function depends on both coordinates X and input features Fin, applied jointly.
    Stated in Section III-C after Eq. (15) and justified only by 'extensive testing', not by a proof.
  • domain assumption The covariance matrix C has a non-degenerate eigen-decomposition so that the frame F(X) is well defined up to sign flips.
    Assumed in Section III-A; degenerate cases (e.g., symmetric point sets) are not discussed and could make the frame discontinuous.

how reviews work

0 comments
Cite this review

Pith. "Pith review of FA-KPConv: Introducing Euclidean Symmetries to KPConv via Frame Averaging." pith.science (2026). https://pith.science/paper/TJK7QH5W

@misc{pith2026250504485,
  author       = {Pith},
  title        = {Pith review of: FA-KPConv: Introducing Euclidean Symmetries to KPConv via Frame Averaging},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TJK7QH5W}},
  note         = {Machine review of arXiv:2505.04485}
}
read the original abstract

We present Frame-Averaging Kernel-Point Convolution (FA-KPConv), a neural network architecture built on top of the well-known KPConv, a widely adopted backbone for 3D point cloud analysis. Even though invariance and/or equivariance to Euclidean transformations are required for many common tasks, KPConv-based networks can only approximately achieve such properties when training on large datasets or with significant data augmentations. Using Frame Averaging, we allow to flexibly customize point cloud neural networks built with KPConv layers, by making them exactly invariant and/or equivariant to translations, rotations and/or reflections of the input point clouds. By simply wrapping around an existing KPConv-based network, FA-KPConv embeds geometrical prior knowledge into it while preserving the number of learnable parameters and not compromising any input information. We showcase the benefit of such an introduced bias for point cloud classification and point cloud registration, especially in challenging cases such as scarce training data or randomly rotated test data.

Figures

Figures reproduced from arXiv: 2505.04485 by the authors.

Figure 1
Figure 1. Diagrams illustrating how (a) the equivariant function [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Qualitative comparison of registration results on rotated 3DLoMatch [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 24 canonical work pages

  1. [1]

    Kpconv: Flexible and deformable convolution for point clouds,

    H. Thomas, C. R. Qi, J.-E. Deschaud, B. Marcotegui, F. Goulette, and L. J. Guibas, “Kpconv: Flexible and deformable convolution for point clouds,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) , October 2019

  2. [2]

    Gfnet: Geometric flow network for 3d point cloud semantic segmentation,

    H. Qiu, B. Yu, and D. Tao, “Gfnet: Geometric flow network for 3d point cloud semantic segmentation,” 2022. [Online]. Available: https://arxiv.org/abs/2207.02605

  3. [3]

    Regtr: End-to-end point cloud correspon- dences with transformers,

    Z. J. Yew and G. H. Lee, “Regtr: End-to-end point cloud correspon- dences with transformers,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , June 2022, pp. 6677–6686

  4. [4]

    A Strong Baseline for Point Cloud Registration via Direct Superpoints Matching

    A. Gupta, Y . Xie, H. Singh, and H. Jiang, “A strong baseline for point cloud registration via direct superpoints matching,” 2024. [Online]. Available: https://arxiv.org/abs/2307.01362

  5. [5]

    ObPose: Leveraging Pose for Object-Centric Scene Inference and Generation in 3D

    Y . Wu, O. P. Jones, and I. Posner, “Obpose: Leveraging pose for object-centric scene inference and generation in 3d,” arXiv preprint arXiv:2206.03591, 2022

  6. [6]

    Geotransformer: Fast and robust point cloud registration with geometric transformer,

    Z. Qin, H. Yu, C. Wang, Y . Guo, Y . Peng, S. Ilic, D. Hu, and K. Xu, “Geotransformer: Fast and robust point cloud registration with geometric transformer,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 8, pp. 9806–9821, 2023

  7. [7]

    Advancements in point cloud data augmentation for deep learning: A survey,

    Q. Zhu, L. Fan, and N. Weng, “Advancements in point cloud data augmentation for deep learning: A survey,” Pattern Recognition , vol. 153, p. 110532, 2024. [Online]. Available: https://www.sciencedirect. com/science/article/pii/S0031320324002838

  8. [8]

    Frame averaging for invariant and equivariant network design,

    O. Puny, M. Atzmon, H. Ben-Hamu, I. Misra, A. Grover, E. J. Smith, and Y . Lipman, “Frame averaging for invariant and equivariant network design,” arXiv preprint arXiv:2110.03336 , 2021

Show all 43 references
  1. [9]

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

    C. R. Qi, H. Su, K. Mo, and L. J. Guibas, “Pointnet: Deep learning on point sets for 3d classification and segmentation,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , July 2017

  2. [10]

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

    C. R. Qi, L. Yi, H. Su, and L. J. Guibas, “Pointnet++: Deep hierarchical feature learning on point sets in a metric space,” Advances in neural information processing systems , vol. 30, 2017

  3. [11]

    Pointnext: Revisiting pointnet++ with improved training and scaling strategies,

    G. Qian, Y . Li, H. Peng, J. Mai, H. Hammoud, M. Elhoseiny, and B. Ghanem, “Pointnext: Revisiting pointnet++ with improved training and scaling strategies,” Advances in neural information processing systems, vol. 35, pp. 23 192–23 204, 2022

  4. [12]

    Spidercnn: Deep learning on point sets with parameterized convolutional filters,

    Y . Xu, T. Fan, M. Xu, L. Zeng, and Y . Qiao, “Spidercnn: Deep learning on point sets with parameterized convolutional filters,” in Proceedings of the European Conference on Computer Vision (ECCV) , September 2018

  5. [13]

    Pointcnn: Convolution on x-transformed points,

    Y . Li, R. Bu, M. Sun, W. Wu, X. Di, and B. Chen, “Pointcnn: Convolution on x-transformed points,” Advances in neural information processing systems, vol. 31, 2018

  6. [14]

    Paconv: Position adaptive convolution with dynamic kernel assembling on point clouds,

    M. Xu, R. Ding, H. Zhao, and X. Qi, “Paconv: Position adaptive convolution with dynamic kernel assembling on point clouds,” in Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2021, pp. 3173–3182

  7. [15]

    Pointconv: Deep convolutional networks on 3d point clouds,

    W. Wu, Z. Qi, and L. Fuxin, “Pointconv: Deep convolutional networks on 3d point clouds,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , June 2019

  8. [16]

    Dynamic graph cnn for learning on point clouds,

    Y . Wang, Y . Sun, Z. Liu, S. E. Sarma, M. M. Bronstein, and J. M. Solomon, “Dynamic graph cnn for learning on point clouds,” ACM Transactions on Graphics (tog) , vol. 38, no. 5, pp. 1–12, 2019

  9. [17]

    Dynamic edge-conditioned filters in convolutional neural networks on graphs,

    M. Simonovsky and N. Komodakis, “Dynamic edge-conditioned filters in convolutional neural networks on graphs,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , July 2017

  10. [18]

    Adaptive graph convolution for point cloud analysis,

    H. Zhou, Y . Feng, M. Fang, M. Wei, J. Qin, and T. Lu, “Adaptive graph convolution for point cloud analysis,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) , October 2021, pp. 4965–4974

  11. [19]

    Pct: Point cloud transformer,

    M.-H. Guo, J.-X. Cai, Z.-N. Liu, T.-J. Mu, R. R. Martin, and S.-M. Hu, “Pct: Point cloud transformer,” Computational Visual Media, vol. 7, pp. 187–199, 2021

  12. [20]

    Point transformer,

    H. Zhao, L. Jiang, J. Jia, P. H. Torr, and V . Koltun, “Point transformer,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), October 2021, pp. 16 259–16 268

  13. [21]

    Point transformer v2: Grouped vector attention and partition-based pooling,

    X. Wu, Y . Lao, L. Jiang, X. Liu, and H. Zhao, “Point transformer v2: Grouped vector attention and partition-based pooling,” Advances in Neural Information Processing Systems , vol. 35, pp. 33 330–33 342, 2022

  14. [22]

    Point transformer v3: Simpler faster stronger,

    X. Wu, L. Jiang, P.-S. Wang, Z. Liu, X. Liu, Y . Qiao, W. Ouyang, T. He, and H. Zhao, “Point transformer v3: Simpler faster stronger,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2024, pp. 4840–4851

  15. [23]

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

    X. Yu, L. Tang, Y . Rao, T. Huang, J. Zhou, and J. Lu, “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 (CVPR), June 2022, pp. 19 313–19 322

  16. [24]

    Masked autoencoders for point cloud self-supervised learning,

    Y . Pang, W. Wang, F. E. Tay, W. Liu, Y . Tian, and L. Yuan, “Masked autoencoders for point cloud self-supervised learning,” in European conference on computer vision . Springer, 2022, pp. 604–621

  17. [25]

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

    R. Zhang, Z. Guo, P. Gao, R. Fang, B. Zhao, D. Wang, Y . Qiao, and H. Li, “Point-m2ae: multi-scale masked autoencoders for hierarchical point cloud pre-training,” Advances in neural information processing systems, vol. 35, pp. 27 061–27 074, 2022

  18. [26]

    Spatial transformer networks,

    M. Jaderberg, K. Simonyan, A. Zisserman et al. , “Spatial transformer networks,” Advances in neural information processing systems , vol. 28, 2015

  19. [27]

    Geometric invariance of pointnet,

    H. Le, “Geometric invariance of pointnet,” B.S. thesis, 2021

  20. [28]

    Group equivariant convolutional networks,

    T. Cohen and M. Welling, “Group equivariant convolutional networks,” in International conference on machine learning . PMLR, 2016, pp. 2990–2999

  21. [29]

    Steerable cnns,

    T. S. Cohen and M. Welling, “Steerable cnns,” arXiv preprint arXiv:1612.08498, 2016

  22. [30]

    Spherical cnns,

    T. S. Cohen, M. Geiger, J. K ¨ohler, and M. Welling, “Spherical cnns,” arXiv preprint arXiv:1801.10130 , 2018

  23. [31]

    Equivariant point network for 3d point cloud analysis,

    H. Chen, S. Liu, W. Chen, H. Li, and R. Hill, “Equivariant point network for 3d point cloud analysis,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2021, pp. 14 514–14 523

  24. [32]

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

    N. Thomas, T. Smidt, S. Kearnes, L. Yang, L. Li, K. Kohlhoff, and P. Riley, “Tensor field networks: Rotation-and translation-equivariant neural networks for 3d point clouds,” arXiv preprint arXiv:1802.08219, 2018

  25. [33]

    Se (3)-transformers: 3d roto-translation equivariant attention networks,

    F. Fuchs, D. Worrall, V . Fischer, and M. Welling, “Se (3)-transformers: 3d roto-translation equivariant attention networks,” Advances in neural information processing systems , vol. 33, pp. 1970–1981, 2020

  26. [34]

    Clusternet: Deep hierarchical cluster network with rigorously rotation-invariant representation for point cloud analysis,

    C. Chen, G. Li, R. Xu, T. Chen, M. Wang, and L. Lin, “Clusternet: Deep hierarchical cluster network with rigorously rotation-invariant representation for point cloud analysis,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 4994– 5002

  27. [35]

    Rotation invariant convolutions for 3d point clouds deep learning,

    Z. Zhang, B.-S. Hua, D. W. Rosen, and S.-K. Yeung, “Rotation invariant convolutions for 3d point clouds deep learning,” in 2019 International conference on 3d vision (3DV) . IEEE, 2019, pp. 204–213

  28. [36]

    Ppfnet: Global context aware local features for robust 3d point matching,

    H. Deng, T. Birdal, and S. Ilic, “Ppfnet: Global context aware local features for robust 3d point matching,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2018, pp. 195– 205

  29. [37]

    Srinet: Learning strictly rotation-invariant representations for point cloud classification and segmentation,

    X. Sun, Z. Lian, and J. Xiao, “Srinet: Learning strictly rotation-invariant representations for point cloud classification and segmentation,” in Proceedings of the 27th ACM international conference on multimedia , 2019, pp. 980–988

  30. [38]

    Darboux frames, snakes, and super-quadrics: Geometry from the bottom up,

    F. P. Ferrie, J. Lagarde, and P. Whaite, “Darboux frames, snakes, and super-quadrics: Geometry from the bottom up,” IEEE transactions on pattern analysis and machine intelligence , vol. 15, no. 8, pp. 771–784, 1993

  31. [39]

    E (n) equivariant graph neural networks,

    V . G. Satorras, E. Hoogeboom, and M. Welling, “E (n) equivariant graph neural networks,” in International conference on machine learning . PMLR, 2021, pp. 9323–9332

  32. [40]

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

    C. Deng, O. Litany, Y . Duan, A. Poulenard, A. Tagliasacchi, and L. J. Guibas, “Vector neurons: A general framework for so (3)-equivariant networks,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2021, pp. 12 200–12 209

  33. [41]

    Boosting deep neural networks with geometrical prior knowledge: A survey,

    M. Rath and A. P. Condurache, “Boosting deep neural networks with geometrical prior knowledge: A survey,” Artificial Intelligence Review , vol. 57, no. 4, p. 95, 2024

  34. [42]

    3d shapenets: A deep representation for volumetric shapes,

    Z. Wu, S. Song, A. Khosla, F. Yu, L. Zhang, X. Tang, and J. Xiao, “3d shapenets: A deep representation for volumetric shapes,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2015, pp. 1912–1920

  35. [43]

    3dmatch: Learning local geometric descriptors from rgb-d reconstruc- tions,

    A. Zeng, S. Song, M. Nießner, M. Fisher, J. Xiao, and T. Funkhouser, “3dmatch: Learning local geometric descriptors from rgb-d reconstruc- tions,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 1802–1811

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.