Rotation Invariant Convolutions for 3D Point Clouds Deep Learning

Binh-Son Hua; David W. Rosen; Sai-Kit Yeung; Zhiyuan Zhang

Rotation Invariant Convolutions for 3D Point Clouds Deep Learning

T0 review · 0 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read Distances and angles to a centroid reference vector make point-cloud convolutions rotation invariant.

desk verdict A clean and honest rotation-invariant point cloud convolution whose central claim survives scrutiny; the soft spots are pseudocode detail and evaluation rigor, not the math. read the letter →

arxiv 1908.06297 v1 pith:FF2EU552 submitted 2019-08-17 cs.CV

Zhiyuan Zhang , Binh-Son Hua , David W. Rosen , Sai-Kit Yeung This is my paper · ORCID

classification cs.CV

keywords rotationinvariantconvolutionpointcloud3DdeeplearningobjectclassificationpartsegmentationModelNet40ShapeNetgeometricfeatures

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 sets out to build a convolution operator for 3D point clouds that is invariant to rigid transformations and to the order in which points are listed, so that a network trained on mildly rotated data still works on arbitrarily rotated data. It claims this can be done directly in Euclidean space by replacing raw coordinates with four low-level geometric measurements—two distances and two angles—taken relative to a reference direction anchored at a point and the centroid of its neighborhood. Because these measurements do not change under rotations or translations of the local set, the resulting operator, RIConv, does not need augmentation with many rotations or a spherical grid. The payoff is consistent accuracy across rotation regimes: the reported classification accuracy on ModelNet40 is 86.5%, 86.4%, and 86.4% for the z/z, SO3/SO3, and z/SO3 settings, with an accuracy standard deviation of 0.1, and part segmentation mIoU of 75.5 and 75.3 on ShapeNet. A sympathetic reader would take this as evidence that simple geometric features, rather than learned canonical orientations, are enough to make point-cloud learning stable to 6DoF transformations.

What carries the argument

The load-bearing object is the rotation-invariant feature $\mathrm{RIF}(x; \vec{pm}) = [d_0, d_1, \alpha_0, \alpha_1]$, computed for every point $x$ in a local neighborhood of a reference point $p$, with $m$ the centroid of that neighborhood. The vector $\vec{pm}$ serves as a local orientation reference: distances and angles to $p$ and $m$ do not change under rigid motions, so the features carry no dependence on the global frame. A shared MLP lifts these features, the local space is divided into ordered bins along $\vec{pm}$, and each bin is max-pooled; the ordered bin features then make a 1D convolution well-defined, which is how point-order invariance is obtained. When $p$ and $m$ coincide—symmetrically distributed neighbors—the paper selects the farthest point to $p$ as $m$ to avoid the degenerate reference vector.

What would settle it

Retrain the RIConv classification network on ModelNet40 with only z-axis rotations, then test it on a fixed set of non-axis-aligned SO3 rotations and measure per-category accuracy; if the overall accuracy falls well below the reported 86.4% or the accuracy standard deviation across repeated rotation sets exceeds roughly 0.1, the claimed rotation invariance is not holding. A second, more local check: place a point at the exact centroid of its neighborhood and compare the bin assignments and output features before and after an infinitesimal rotation; any abrupt change would reveal the degenerate-reference failure.

Watch

Extended reading notes

Core claim

The central discovery is that a point cloud convolution can be made rotation invariant without leaving Euclidean space, by defining each local point's feature as $[d_0, d_1, \alpha_0, \alpha_1]$: the distances from the point to the reference point and to the neighborhood centroid, and the angles those two segments make with the reference vector pointing from the reference point to the centroid. The operator then bins the lifted features along that reference vector, max-pools per bin, and applies a 1D convolution, which absorbs the issue of point ordering. The paper's claim is that this single operator is agnostic to both point rotation and point order, and that a compact network built from it produces nearly identical accuracy whether it is trained and tested with z-axis rotations, trained and tested with arbitrary SO3 rotations, or trained with z-rotations and tested with SO3 rotations. In the hardest case, training with z-rotations and testing with SO3 rotations, the network keeps 86.4% accuracy on ModelNet40 classification, where PointNet, PointNet++, and PointCNN drop to 12.8%, 28.6%, and 41.2%, respectively; on ShapeNet part segmentation it reaches 75.3 mIoU in the same cross-rotation setting.

Load-bearing premise

The whole construction depends on the centroid-to-point vector staying a well-defined local reference under rotation, and on the farthest-point fallback, used when that vector collapses to zero, not introducing discontinuities.

Editorial extensions

If this is right

A network trained only with z-axis rotations can be deployed on arbitrary SO3-rotated point clouds without retraining or rotation augmentation, holding about 86.4% classification accuracy on ModelNet40.
The same operator serves both classification and part segmentation: on ShapeNet part segmentation it reaches 75.5 mIoU under SO3/SO3 and 75.3 under z/SO3, with consistent per-category results.
Rotation invariance is achieved with only 0.70M parameters, roughly five times fewer than PointNet, because the network does not need to memorize many rotated views.
The binning strategy subsumes the point-ordering problem: the ordered, max-pooled bins give a fixed-size feature sequence on which a 1D convolution can act.
Using original coordinates instead of geometric features raises classification accuracy to 91.8% in the no-rotation setting, indicating a trade-off between invariance and discriminative power that the paper leaves open.

Reading between the lines

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

Editorial inference: fusing raw coordinates with the geometric features—rather than dropping coordinates—is the natural next experiment suggested by the paper's own 91.8% coordinate-only result; it could recover discriminative power while preserving cross-rotation consistency.
Editorial inference: the degeneracy noted when the reference point equals the centroid is described in the text but not in Algorithm 1; the farthest-neighbor fallback may preserve invariance for symmetric neighborhoods but could introduce discontinuous bin assignments under small perturbations, so a production implementation should handle that case explicitly.
Editorial inference: the binning discretization fixes an angular resolution; shapes with fine local detail would likely benefit from more bins in the first layer, and this is testable by sweeping the 4/2/1 bin schedule.
Editorial inference: the same recipe of distances and angles could plausibly extend to scale invariance by normalizing the distances, or to non-rigid settings by replacing the centroid with a more stable anchor, though the paper does not explore these options.

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.

Desk Editor's Note

read the letter

Read this one if you care about rotation robustness in point cloud networks. The paper builds a convolution operator that is rotation invariant by construction: for each local neighborhood it defines a reference vector from the point to the centroid, encodes every neighbor by distances and angles to that vector, bins the points along the vector, and runs a 1D convolution over the bin features. That is the whole approach, and it works as claimed.

What is new is the particular combination. Distances and angles as local descriptors are old, and the binning idea comes from the authors' own ShellNet, but turning those into a point-cloud convolution layer that is provably invariant to rigid motions and point order is a solid, useful contribution. The network is compact (0.7M parameters) and the key result is the consistency: 86.5/86.4/86.4 accuracy across z/z, SO3/SO3, and the hard z/SO3 transfer, with a standard deviation of 0.1. The part segmentation numbers show the same pattern, and the paper is honest that on axis-aligned data it gives up a few points to methods that keep the original coordinates.

Soft spots are minor. Algorithm 1 does not show the degenerate-centroid fallback that Section 3.1 describes in text; when the centroid coincides with the reference point, the pseudocode would produce a zero reference vector. The fallback (use the farthest point) is rotation equivariant, so the invariance claim holds, but the implementation detail should be explicit and the potential feature discontinuity at that transition is worth discussing. Evaluation is single-run, with no error bars; given the small accuracy spread, repeated runs would strengthen the claim. The baseline comparison protocol is also a little underspecified: some numbers appear to come from prior papers, and it is not clear whether all baselines were retrained under identical rotation augmentation for the z/SO3 condition. All of these are addressable.

The citation pattern is fine; the ShellNet self-citation is appropriate because the binning is genuinely from that paper. No missing related work jumps out, and the paper's own limitation paragraph correctly frames the accuracy trade-off.

Verdict: the central claim, exact rotation invariance plus empirical consistency, holds up. I would send this to peer review; it deserves a careful referee, and the fixes are straightforward. For a reading group on rotation-invariant learning, it is worth an hour.

Referee Report

0 major / 5 minor

Summary. The paper proposes RIConv, a convolution operator for 3D point clouds that is invariant to rigid motions and to the ordering of input points. For each reference point p, the K nearest neighbors form a local point set; the centroid m is computed, and the vector pm serves as a local reference direction. Each neighbor x is encoded by four scalar features, two distances and two angles (Eq. 1), which are lifted by a shared MLP. The local points are then divided into ordered bins along pm, max-pooled per bin, and processed by a 1D convolution to produce the output feature. The operator is used in a compact CNN for object classification and in an encoder-decoder with skip connections for part segmentation. Experiments on ModelNet40 report 86.5%, 86.4%, and 86.4% classification accuracy for the z/z, SO3/SO3, and z/SO3 rotation settings, and on ShapeNet part segmentation report 75.5 and 75.3 mIoU for SO3/SO3 and z/SO3. The paper also includes ablations, per-category tables, and a discussion of limitations.

Significance. If the results hold, the paper makes a useful contribution: a simple and compact point-cloud convolution whose rotation invariance is guaranteed by construction rather than achieved through data augmentation. The core idea is clean and well grounded: the features in Eq. (1) are distances and angles measured relative to pm, so rigid transformations leave them unchanged, and the binning plus max-pooling step makes the operator symmetric with respect to the order of neighboring points. The reported consistency between SO3/SO3 and z/SO3 (86.4 vs 86.4 for classification, 75.5 vs 75.3 for part segmentation) is strong empirical evidence for the invariance claim. The manuscript also provides public code, ablation studies, per-category results, and an honest discussion of the accuracy trade-off in the z/z setting. These are concrete strengths that support the central claim.

minor comments (5)

[Section 3.1 / Algorithm 1] The degenerate-centroid fallback described in Section 3.1 is not present in Algorithm 1, which unconditionally computes pm as m - p. As written, the pseudocode produces a zero reference vector whenever p and m coincide, leaving the angle features and bin boundaries undefined; please add the farthest-point fallback to the algorithm and specify how binning is defined in that case. The fallback is rotation-equivariant, so the central invariance claim is unaffected, but the current pseudocode is not a complete specification of the operator.
[Table 1 / Section 5.1] The column labeled "Acc. std." is described in the text as the standard deviation of accuracy across the three rotation scenarios, not as a repeated-run or seed-to-seed variation. Please state this explicitly in the table caption to avoid misinterpretation, and consider reporting repeated-run statistics for at least the main configurations.
[Algorithm 1, line 6] The binning step is described only verbally as "divide local space into s bins along pm"; please provide the explicit projection formula (for example, t = ((x - p) · pm) / ||pm||^2) and state how empty bins are handled during max-pooling.
[Equation (1)] The symbols d0, d1, α0, and α1 are not formally defined in the text; please define them explicitly as the distances from x to p and m and the angles between (x-p) or (x-m) and the reference vector pm, respectively.
[Section 3.2 / point-order claim] The paper states that the network is agnostic to point orders, yet nearest-neighbor search and farthest-point sampling can have order-dependent tie-breaking; a sentence clarifying that ties are resolved consistently or by input index would make the claim precise.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: RIConv's rotation invariance is a direct geometric construction, and the reported accuracies are independent empirical measurements.

full rationale

The central derivation is self-contained. Equation (1) defines RIF(x; pm) = [d0,d1,alpha0,alpha1] as Euclidean distances and angles measured against pm = m - p, and the invariance claim follows directly from the fact that rigid transformations preserve distances and angles while sending p, m, and every neighbor to correspondingly rotated positions; the features and the binning cells therefore rotate with the cloud. This is a construction, not a reduction of a prediction to its input. The binning and maxpooling mechanism in Section 3.2 and Algorithm 1 is fully specified and gives point-order agnosticism through the permutation invariance of maxpooling; the reference to ShellNet [45] is only a relevance note and is not load-bearing, and [44] is a self-citation of the conference version, not an evidentiary premise. The benchmark accuracies (86.5/86.4/86.4 and 75.5/75.3 mIoU) are measured outcomes, not fitted constants. The degenerate-centroid caveat in Section 3.1 (farthest-point fallback) is under-specified in Algorithm 1 and can produce discontinuous features when the centroid crosses the reference point, but it preserves rotation equivariance and therefore does not threaten the central claim; this is a robustness and reproducibility issue, not circularity.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical or mathematical entities. Its central claim rests on standard Euclidean geometry, the equivariance of centroid computation, and the empirical assumption that simple distance and angle features are discriminative enough for the tested tasks. The free parameters are architecture hyperparameters that affect accuracy but not the rotation invariance guarantee.

free parameters (5)

Number of bins per convolution layer = 4, 2, 1
Manually chosen in Section 4 ('We empirically set the number of bins...') to balance accuracy and speed; accuracy changes with this parameter, but the rotation invariance property does not depend on the exact values.
Neighborhood size K = 64, 32, 16
Set for the three convolution layers; controls local context; chosen empirically to capture point distribution.
Number of output feature vectors for classifier = 64
Ablation (Table 4) shows 64 vectors give 86.5% vs 84.8% for a single vector; not essential to the invariance property.
Downsampling point counts = 256, 128, 64 (classification); 512, 128, 32 (segmentation)
Architecture choices for hierarchical feature extraction, reported in Section 4.
Number of convolution layers = 3
Ablation shows three layers balance accuracy and time; one layer drops accuracy to 46.8%.

assumptions (4)

standard math Euclidean distances and angles are invariant under rigid transformations (translations and rotations).
Foundation of the RIF features in Section 3.1.
standard math The centroid of a point set transforms equivariantly under rotations: if all points rotate by R, the centroid rotates to the rotated centroid.
Used in the definition of reference vector pm = m - p in Section 3.1 and Algorithm 1.
standard math Farthest point sampling is equivariant under rigid transformations, so representative points rotate consistently with the cloud.
Used for representative point selection in Section 3.2.
domain assumption Local rotation-invariant geometric features (distances and angles) retain enough discriminative information for object classification and part segmentation.
The paper's accuracy results rely on this empirical premise; not proven but supported by ablations in Section 5.3.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Rotation Invariant Convolutions for 3D Point Clouds Deep Learning." pith.science (2026). https://pith.science/paper/FF2EU552

@misc{pith2026190806297,
  author       = {Pith},
  title        = {Pith review of: Rotation Invariant Convolutions for 3D Point Clouds Deep Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FF2EU552}},
  note         = {Machine review of arXiv:1908.06297}
}

read the original abstract

Recent progresses in 3D deep learning has shown that it is possible to design special convolution operators to consume point cloud data. However, a typical drawback is that rotation invariance is often not guaranteed, resulting in networks being trained with data augmented with rotations. In this paper, we introduce a novel convolution operator for point clouds that achieves rotation invariance. Our core idea is to use low-level rotation invariant geometric features such as distances and angles to design a convolution operator for point cloud learning. The well-known point ordering problem is also addressed by a binning approach seamlessly built into the convolution. This convolution operator then serves as the basic building block of a neural network that is robust to point clouds under 6DoF transformations such as translation and rotation. Our experiment shows that our method performs with high accuracy in common scene understanding tasks such as object classification and segmentation. Compared to previous works, most importantly, our method is able to generalize and achieve consistent results across different scenarios in which training and testing can contain arbitrary rotations.

Figures

Figures reproduced from arXiv: 1908.06297 by the authors.

**Figure 2.** RIConv operator construction. (a) For an input point cloud with/without associated features, representative points (red dots) are [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Our proposed network architecture. We follow a convolutional neural network design for classification and part segmentation. Skip [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: Qualitative results of object part segmentation task with z/SO3 scenario. Our method has the state-of-the-art performance while [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

**Figure 1.** Figure 1: Overall accuracy vs. epochs plot of object classifica [PITH_FULL_IMAGE:figures/full_fig_p011_1.png]

**Figure 2.** Figure 2: mIoU vs. epochs plot of object part segmentation with [PITH_FULL_IMAGE:figures/full_fig_p011_2.png]

**Figure 3.** Figure 3: Additional qualitative results of object part segmentation task with z/SO3 scenario. Our method has the state-of-the-art performance [PITH_FULL_IMAGE:figures/full_fig_p013_3.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

45 extracted references · 43 canonical work pages

[1]

write newline

" write newline "" before.all 'output.state := FUNCTION fin.entry add.period write newline FUNCTION new.block output.state before.all = 'skip after.block 'output.state := if FUNCTION new.sentence output.state after.block = 'skip output.state before.all = 'skip after.sentence 'output.state := if if FUNCTION not #0 #1 if FUNCTION and 'skip pop #0 if FUNCTIO...
[2]

Abadi, P

M. Abadi, P. Barham, J. Chen, Z. Chen, A. Davis, J. Dean, M. Devin, S. Ghemawat, G. Irving, M. Isard, et al. Tensorflow: A system for large-scale machine learning. In 12th \ USENIX \ Symposium on Operating Systems Design and Implementation ( \ OSDI \ 16) , pages 265--283, 2016

work page 2016
[3]

Armeni, O

I. Armeni, O. Sener, A. R. Zamir, H. Jiang, I. Brilakis, M. Fischer, and S. Savarese. 3d semantic parsing of large-scale indoor spaces. In CVPR , 2016

work page 2016
[4]

Ben-Shabat, M

Y. Ben-Shabat, M. Lindenbaum, and A. Fischer. 3dmfv: Three-dimensional point cloud classification in real-time using convolutional neural networks. IEEE Robotics and Automation Letters , 3:3145--3152, 2018

work page 2018
[5]

M. M. Bronstein , J. Bruna , Y. LeCun , A. Szlam , and P. Vandergheynst . Geometric deep learning: Going beyond euclidean data. IEEE Signal Processing Magazine , 34(4):18--42, 2017

work page 2017
[6]

A. X. Chang, T. A. Funkhouser, L. J. Guibas, P. Hanrahan, Q.-X. Huang, Z. Li, S. Savarese, M. Savva, S. Song, H. Su, J. Xiao, L. Yi, and F. Yu. Shapenet: An information-rich 3d model repository. arXiv preprint arXiv:1512.03012 , 2015

arXiv 2015
[7]

A. Dai, A. X. Chang, M. Savva, M. Halber, T. Funkhouser, and M. Niessner. Scannet: Richly-annotated 3d reconstructions of indoor scenes. In CVPR , pages 5828--5839, 2017

work page 2017
[8]

H. Deng, T. Birdal, and S. Ilic. Ppf-foldnet: Unsupervised learning of rotation invariant 3D local descriptors. In ECCV , 2018

work page 2018

Show all 45 references

[9]

Esteves, C

C. Esteves, C. Allen-Blanchette, A. Makadia, and K. Daniilidis. Learning so (3) equivariant representations with spherical cnns. In ECCV , pages 52--68, 2018

2018
[10]

Groueix, M

T. Groueix, M. Fisher, V. G. Kim, B. Russell, and M. Aubry. AtlasNet: A Papier-M\^ach\'e Approach to Learning 3D Surface Generation . In CVPR , 2018

2018
[11]

Hermosilla, T

P. Hermosilla, T. Ritschel, P.-P. Vazquez, A. Vinacua, and T. Ropinski. Monte carlo convolution for learning on non-uniformly sampled point clouds. ACM Transactions on Graphics , 2018

2018
[12]

Hua, Q.-H

B.-S. Hua, Q.-H. Pham, D. T. Nguyen, M.-K. Tran, L.-F. Yu, and S.-K. Yeung. Scenenn: A scene meshes dataset with annotations. In International Conference on 3D Vision , 2016

2016
[13]

Hua, M.-K

B.-S. Hua, M.-K. Tran, and S.-K. Yeung. Point-wise convolutional neural network. In CVPR , 2018

2018
[14]

Huang, W

Q. Huang, W. Wang, and U. Neumann. Recurrent slice networks for 3d segmentation on point clouds. In CVPR , 2018

2018
[15]

Khoury, Q.-Y

M. Khoury, Q.-Y. Zhou, and V. Koltun. Learning compact geometric features. In ICCV , 2017

2017
[16]

Klokov and V

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

2017
[17]

Landrieu and M

L. Landrieu and M. Simonovsky. Large-scale point cloud semantic segmentation with superpoint graphs. In CVPR , 2018

2018
[18]

Le and Y

T. Le and Y. Duan. Pointgrid: A deep network for 3d shape understanding. CVPR , 2018

2018
[19]

J. Li, B. M. Chen, and G. H. Lee. So-net: Self-organizing network for point cloud analysis. In CVPR , 2018

2018
[20]

Y. Li, R. Bu, M. Sun, and B. Chen. Pointcnn: Convolution on x-transformed points. Advances in Neural Information Processing Systems , 2018

2018
[21]

Y. Li, S. Pirk, H. Su, C. R. Qi, and L. J. Guibas. Fpnn: Field probing neural networks for 3d data. In Advances in Neural Information Processing Systems , pages 307--315, 2016

2016
[22]

X. Liu, Z. Han, Y.-S. Liu, and M. Zwicker. Point2sequence: Learning the shape representation of 3d point clouds with an attention-based sequence to sequence network. In Association for the Advancement of Artificial Intelligence , 2019

2019
[23]

Masci, D

J. Masci, D. Boscaini, M. M. Bronstein, and P. Vandergheynst. Geodesic convolutional neural networks on riemannian manifolds. In The IEEE International Conference on Computer Vision (ICCV) Workshops , 2015

2015
[24]

Maturana and S

D. Maturana and S. Scherer. VoxNet: A 3D Convolutional Neural Network for Real-Time Object Recognition . In International Conference on Intelligent Robots and Systems , 2015

2015
[25]

C. R. Qi, H. Su, K. Mo, and L. J. Guibas. Pointnet: Deep learning on point sets for 3d classification and segmentation. In CVPR , 2017

2017
[26]

C. R. Qi, H. Su, M. Nie ner, A. Dai, M. Yan, and L. J. Guibas. Volumetric and multi-view cnns for object classification on 3d data. In CVPR , pages 5648--5656, 2016

2016
[27]

C. R. Qi, L. Yi, H. Su, and L. J. Guibas. Pointnet++: Deep hierarchical feature learning on point sets in a metric space. In Advances in Neural Information Processing Systems , pages 5105--5114, 2017

2017
[28]

Y. Rao, J. Lu, and J. Zhou. Spherical fractal convolutional neural networks for point cloud recognition. In Computer Vision and Pattern Recognition (CVPR) , 2019

2019
[29]

Riegler, A

G. Riegler, A. O. Ulusoy, and A. Geiger. Octnet: Learning deep 3d representations at high resolutions. In CVPR , 2017

2017
[30]

Ronneberger, P

O. Ronneberger, P. Fischer, and T. Brox. U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention , pages 234--241. Springer, 2015

2015
[31]

Y. Shen, C. Feng, Y. Yang, and D. Tian. Mining point cloud local structures by kernel correlation and graph pooling. In CVPR , volume 4, 2018

2018
[32]

H. Su, S. Maji, E. Kalogerakis, and E. Learned-Miller. Multi-view convolutional neural networks for 3d shape recognition. In International Conference on Computer Vision , pages 945--953, 2015

2015
[33]

Tatarchenko, J

M. Tatarchenko, J. Park, V. Koltun, and Q.-Y. Zhou. Tangent convolutions for dense prediction in 3d. In CVPR , pages 3887--3896, 2018

2018
[34]

P.-S. Wang, Y. Liu, Y.-X. Guo, C.-Y. Sun, and X. Tong. O-cnn: Octree-based convolutional neural networks for 3d shape analysis. ACM Transactions on Graphics , 36(4):72, 2017

2017
[35]

Wang, C.-Y

P.-S. Wang, C.-Y. Sun, Y. Liu, and X. Tong. Adaptive o-cnn: A patch-based deep representation of 3d shapes. ACM Transactions on Graphics , 2018

2018
[36]

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 , 2019

2019
[37]

Z. Wu, S. Song, A. Khosla, F. Yu, L. Zhang, X. Tang, and J. Xiao. 3d shapenets: A deep representation for volumetric shapes. In CVPR , pages 1912--1920, 2015

1912
[38]

S. Xie, S. Liu, Z. Chen, and Z. Tu. Attentional shapecontextnet for point cloud recognition. In CVPR , 2018

2018
[39]

Y. Xu, T. Fan, M. Xu, L. Zeng, and Y. Qiao. Spidercnn: Deep learning on point sets with parameterized convolutional filters. In ECCV , 2018

2018
[40]

Y. Yang, C. Feng, Y. Shen, and D. Tian. Foldingnet: Point cloud auto-encoder via deep grid deformation. In CVPR , 2018

2018
[41]

L. Yi, V. G. Kim, D. Ceylan, I.-C. Shen, M. Yan, H. Su, C. Lu, Q. Huang, A. Sheffer, and L. Guibas. A scalable active framework for region annotation in 3d shape collections. ACM Transactions on Graphics , 2016

2016
[42]

L. Yi, H. Su, X. Guo, and L. J. Guibas. Syncspeccnn: Synchronized spectral cnn for 3d shape segmentation. In CVPR , pages 2282--2290, 2017

2017
[43]

A. Zeng, S. Song, M. Nie ner, M. Fisher, J. Xiao, and T. Funkhouser. 3DMatch : Learning local geometric descriptors from RGB-D reconstructions. In CVPR , 2017

2017
[44]

Zhang, B.-S

Z. Zhang, B.-S. Hua, D. W. Rosen, and S.-K. Yeung. Rotation invariant convolutions for 3d point clouds deep learning. In International Conference on 3D Vision (3DV) , 2019

2019
[45]

Zhang, B.-S

Z. Zhang, B.-S. Hua, and S.-K. Yeung. Shellnet: Efficient point cloud convolutional neural networks using concentric shells statistics. In International Conference on Computer Vision (ICCV) , 2019

2019

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

write newline

" write newline "" before.all 'output.state := FUNCTION fin.entry add.period write newline FUNCTION new.block output.state before.all = 'skip after.block 'output.state := if FUNCTION new.sentence output.state after.block = 'skip output.state before.all = 'skip after.sentence 'output.state := if if FUNCTION not #0 #1 if FUNCTION and 'skip pop #0 if FUNCTIO...

[2] [2]

Abadi, P

M. Abadi, P. Barham, J. Chen, Z. Chen, A. Davis, J. Dean, M. Devin, S. Ghemawat, G. Irving, M. Isard, et al. Tensorflow: A system for large-scale machine learning. In 12th \ USENIX \ Symposium on Operating Systems Design and Implementation ( \ OSDI \ 16) , pages 265--283, 2016

work page 2016

[3] [3]

Armeni, O

I. Armeni, O. Sener, A. R. Zamir, H. Jiang, I. Brilakis, M. Fischer, and S. Savarese. 3d semantic parsing of large-scale indoor spaces. In CVPR , 2016

work page 2016

[4] [4]

Ben-Shabat, M

Y. Ben-Shabat, M. Lindenbaum, and A. Fischer. 3dmfv: Three-dimensional point cloud classification in real-time using convolutional neural networks. IEEE Robotics and Automation Letters , 3:3145--3152, 2018

work page 2018

[5] [5]

M. M. Bronstein , J. Bruna , Y. LeCun , A. Szlam , and P. Vandergheynst . Geometric deep learning: Going beyond euclidean data. IEEE Signal Processing Magazine , 34(4):18--42, 2017

work page 2017

[6] [6]

A. X. Chang, T. A. Funkhouser, L. J. Guibas, P. Hanrahan, Q.-X. Huang, Z. Li, S. Savarese, M. Savva, S. Song, H. Su, J. Xiao, L. Yi, and F. Yu. Shapenet: An information-rich 3d model repository. arXiv preprint arXiv:1512.03012 , 2015

arXiv 2015

[7] [7]

A. Dai, A. X. Chang, M. Savva, M. Halber, T. Funkhouser, and M. Niessner. Scannet: Richly-annotated 3d reconstructions of indoor scenes. In CVPR , pages 5828--5839, 2017

work page 2017

[8] [8]

H. Deng, T. Birdal, and S. Ilic. Ppf-foldnet: Unsupervised learning of rotation invariant 3D local descriptors. In ECCV , 2018

work page 2018