Pith. sign in

REVIEW 3 major objections 5 minor 39 references

Point Cloud Super Resolution with Adversarial Residual Graph Networks

T0 review · 3 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 AR-GCN, a residual graph network trained with a graph adversarial loss, outperforms existing point cloud upsampling methods on seen and unseen benchmarks.

desk verdict Solid empirical point-cloud upsampling paper with a genuinely wrong justification for its one-sided loss; worth reviewing, but the loss rationale needs a fix. read the letter →

arxiv 1908.02111 v1 pith:M23ZTONW submitted 2019-08-06 cs.GR eess.IV

Huikai Wu , Junge Zhang , Kaiqi Huang This is my paper · ORCID
classification cs.GReess.IV
keywords pointcloudsuper-resolutiongraphconvolutionnetworkadversariallossresiduallearningupsamplingChamferdistanceLS-GAN3Dreconstruction
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 tries to establish that a graph-convolution generator with residual connections and a graph-based adversarial discriminator can upsample a low-resolution point cloud into a high-resolution one better than existing optimization and learning methods. The authors argue that predicting residual displacements between input and output, rather than regressing coordinates directly, exploits the analogy between low- and high-resolution shapes and stabilizes training. They replace hand-designed uniformity losses with a learned graph adversarial loss, claiming it captures properties like continuity automatically. On the PU-Net train-test data, AR-GCN raises F-score from 43.24% to 70.28% and lowers Chamfer distance, and on unseen SHREC15 it improves F-score to 93.07%, which is the basis for the generalization claim.

What carries the argument

The load-bearing object is the residual graph convolution block built on G-conv, where each vertex feature is updated as $f_{l+1}^p = w_0 f_l^p + w_1 \sum_{q \in N(p)} f_l^q$, with neighbors defined by k-nearest neighbors in Euclidean space ($k=8$). Around this, the unpooling block reshapes a G-conv output into $\hat{n} \times 2 \times 3$ displacements $\delta x$ and adds them to the input to double the point count, so the generator predicts residuals rather than absolute coordinates, aided by a skip connection between input and output. The graph discriminator downsamples via farthest point sampling and max-pooling to output 64 patch scores, enabling a graph patch GAN. The training signal is carried by the loss $L = \lambda L_{\mathrm{cd}} + L_G$, with $L_{\mathrm{cd}}$ the one-sided Chamfer distance and $L_G = \|1 - D(\hat{y})\|^2$.

What would settle it

Train AR-GCN on the PU-Net data with the full symmetric Chamfer distance, keeping $\hat{L}_{\mathrm{cd}}$ alongside $L_{\mathrm{cd}}$ and the adversarial term, and evaluate F-score and Chamfer distance on the same test split. If the full-loss model matches or exceeds AR-GCN, the paper's stated reason for dropping the symmetric term is falsified; a second check is counting near-coincident output points in AR-GCN's predictions, since the dropped term was claimed to prevent duplication.

Watch

Extended reading notes

Core claim

The central claim is that AR-GCN achieves state-of-the-art point cloud super-resolution by combining three components: a residual graph convolution generator that upsamples progressively by 2x and predicts per-point offset displacements through unpooling blocks, a graph patch discriminator that classifies local patches of point clouds as real or fake, and a training loss that keeps only the one-sided Chamfer term $L_{\mathrm{cd}}$ plus an LS-GAN adversarial loss $L_G$. The authors hold that the symmetric Chamfer term $\hat{L}_{\mathrm{cd}}$ encourages output points to coincide with the sparse input, causing duplication, so it is removed. Under this design, the reported tables show consistent gains over PU-Net and MLS in Chamfer distance, Earth Mover's distance, F-score, normalized uniformity coefficient, and deviation on both the test split and SHREC15, with roughly equal parameter count.

Load-bearing premise

The paper's results depend on the claim that discarding the symmetric half of the Chamfer distance, keeping only the distance from ground-truth points to output points, gives a sound training signal, because the dropped half is said to cause duplicate points.

Editorial extensions

If this is right

  • If AR-GCN's reported numbers hold, learning-based point cloud upsampling can beat classical MLS on fidelity and uniformity simultaneously, something PU-Net did not achieve.
  • The same architecture upsamples iteratively: applying the 4x model twice turns 1,024 points into roughly 16,384 points while retaining shape details, so fixed-ratio training transfers to larger ratios.
  • Upsampling sparse inputs with AR-GCN recovers classification accuracy: PointNet++ on 1,024 upsampled-from-256 points reaches 79.34% versus 46.96% on the raw 256 points.
  • The graph adversarial loss appears to enforce uniformity without an explicit repulsion term, since normalized uniformity improves despite no hand-designed uniformity loss.
  • AR-GCN degrades gracefully under Gaussian noise and non-uniform sampling, still beating clean-input PU-Net on several metrics.

Reading between the lines

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

  • If the one-sided Chamfer term is the active ingredient, its removal may trade completeness for fidelity; a direct test would measure precision and recall separately on the output points, since the dropped term is what penalizes output points far from any ground-truth point.
  • The residual displacement formulation transfers naturally to point cloud completion and denoising, where input and output share a surface but differ in density and noise; the paper does not claim this extension.
  • The graph patch discriminator could be paired with other generators or used as a trainable loss for mesh or occupancy upsampling, but that is an extension beyond the paper.
  • A practical variant would retain a down-weighted symmetric Chamfer term and check whether F-score changes; the paper's ablation only removes it entirely.
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 / 5 minor

Summary. The paper proposes AR-GCN, a learning-based point cloud super-resolution method combining a residual graph convolutional generator, a graph-based LS-GAN discriminator, and a loss that weighs a one-sided Chamfer distance term against a graph adversarial term. The generator upsamples progressively by a factor of two using graph convolution blocks, unpooling blocks that predict coordinate residuals, and a skip connection from input to output. The method is trained on the PU-Net dataset and evaluated on both the PU-Net test split and the unseen SHREC15 dataset, reporting improvements over PU-Net and MLS on CD, EMD, F-score, NUC, and Deviation. Additional experiments demonstrate robustness to noisy and uneven inputs, iterative upsampling, mesh reconstruction, and a classification application on ModelNet40.

Significance. If the empirical claims hold, AR-GCN is a practical contribution to point cloud upsampling: it combines residual graph convolutions with an adversarial loss, and the held-out SHREC15 results support a generalization claim. The paper ships a public code repository, provides architecture ablations, and reports downstream applications. However, the central empirical claim rests on a training objective whose geometric rationale is stated incorrectly in Section 3.2, and the loss-balance hyperparameter lambda is never reported. These issues are fixable but require additional experiments and reporting before the benchmark conclusions can be accepted as stated.

major comments (3)
  1. [Section 3.2, Eq. (3)] The statement that L_hat_cd 'encourages y_hat to be identical to the LR input, which leads to duplication points' is incorrect. Eq. (3) sums over output points q the squared distance to the nearest point of the ground truth y, so it penalizes output points that are far from the GT surface; it does not pull output points toward the sparse LR input, and duplicates lying on the GT surface would not be penalized. Because the one-sided term Lcd in Eq. (2) only requires every GT point to have a nearby output point, output points that are not the nearest neighbor of any GT point receive no geometric gradient from Lcd. The paper provides no ablation comparing the proposed one-sided objective with the full symmetric Chamfer distance, so it is not possible to tell whether the gains in Tables 1 and 2 come from the residual graph architecture and adversarial loss or from an under-constrained geometric loss. Please correct the justification and add such an ablation with lambda reported.
  2. [Section 4.2, Eq. (1)] Equation (1) defines L(x,y) = lambda Lcd + LG, but the value of lambda is never reported anywhere in Section 4.2 or in the ablation study. Since the two-step training procedure (80 epochs with Lcd, then 40 epochs with the full loss) interacts with this weighting, the loss-balance hyperparameter is essential for reproducibility and for interpreting the relative contribution of the adversarial term. Please report the value and, ideally, a small sensitivity study.
  3. [Tables 1 and 2] The headline comparisons are reported from a single run without error bars or significance tests. While the F-score margins are large (70.28% vs. 43.24% on the train-test split and 93.07% vs. 56.39% on SHREC15), some of the claimed advantages, such as EMD on SHREC15 (0.0031 vs. 0.0032 for MLS) and several NUC entries, are small. Please report mean and standard deviation over multiple training runs, or provide a significance test for the close comparisons, to support the state-of-the-art claim.
minor comments (5)
  1. [Section 3.2, Eq. (2)] Calling Lcd a 'point-wise distance' is misleading: Eq. (2) is a one-sided coverage term from the ground truth to the output, not a per-point correspondence distance. Please rephrase to avoid confusion with the later claim that Lcd measures point-wise distance.
  2. [Section 4.3] The metric column 'NUC with different p' is not explained in the text; please define p, state the range used, and clarify that smaller values are better for both NUC and Deviation.
  3. [Table 3] The row label 'GCN4×points' is visually ambiguous; consider writing 'GCN-4x (points)' or adding a footnote that clarifies the subscript.
  4. [Introduction and Related Work] The abstract claims state-of-the-art performance, but the comparison set contains only PU-Net and MLS. Please either broaden the comparison or explicitly narrow the claim to these baselines.
  5. [Section 4.5] The sentence 'the F-score decreases by about 8% as shown by ResGCN + Lpu and ResGCN' should state whether the change is relative or in percentage points, since the table shows a drop from 68.75% to 60.71%.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the state-of-the-art claim is an empirical benchmark result trained and evaluated on external data.

full rationale

The paper's central claim is an empirical performance claim evaluated on external benchmarks (the PU-Net train-test split and the unseen SHREC15 dataset) with metrics that are not defined in terms of the trained parameters or the training objective. The training loss in Eq. 1, L(x,y) = lambda * Lcd(G(x),y) + LG(G(x)), is used to fit the generator and discriminator, but the reported CD, EMD, F-score, NUC, and Deviation values are measured after training on held-out or unseen point clouds, so the predictions do not reduce by construction to the fitted loss. The ablation study in Table 3 provides an internal comparison of architecture variants and training strategies, and the comparisons against MLS and PU-Net are external baselines. The paper's justification for dropping the symmetric Chamfer term L_hat_cd is questionable: Eq. 3 sums over output points the squared distance to the nearest ground-truth point, which penalizes output points far from the ground truth rather than encouraging identity with the LR input. However, an incorrect or weakly supported design justification is not circular reasoning, because the benchmark results remain externally falsifiable and the model is not defined in terms of its reported metrics. The unreported value of lambda and the hand-selected two-stage training schedule are reproducibility concerns, not circular dependency. There are no load-bearing self-citations by the present authors; the cited works on PU-Net, LS-GAN, GCN, and patch-GAN are prior external references used for comparison or building blocks, not as evidence that reduces the central claim to its own inputs. Overall, no equation or experimental result in the paper is equivalent to its own input by definition, and no fitted parameter is renamed as a prediction.

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

The paper introduces no new physical entities; its graph discriminator and residual graph convolution blocks are software components. The central claim rests on hand-selected hyperparameters, an unreported loss weight, and dataset-specific assumptions about what makes a good point cloud.

free parameters (5)
  • lambda (loss balance) = not reported
    The joint loss in Eq. 1 combines Chamfer distance and adversarial loss with weight lambda, but the paper never reports its value. The central training objective is therefore under-specified.
  • k (nearest neighbors) = 8
    The graph connectivity for the generator and discriminator feature nets and G-conv layers is chosen by hand in Sections 3.3.1, 3.3.3, and 3.4.
  • channels and depth = 128 channels and 12 residual layers in the generator; 64 channels and 4 layers in the discriminator
    Architecture capacity is chosen by hand and directly affects the quality of the generated point cloud.
  • training schedule = 80 epochs with Lcd, then 40 epochs with joint loss; Adam at learning rate 0.001, batch size 28
    The two-stage training schedule matters; the ablation shows AR-GCN without fine-tuning is worse, so these choices affect the central claim.
  • discriminator output patch size = 64 points
    The graph patch GAN output size is chosen by hand in Section 3.4 and determines how local the realism constraint is.
assumptions (5)
  • domain assumption A k-NN graph in Euclidean space is an adequate local structure representation for point cloud generation.
    The generator and discriminator define neighborhoods as the k nearest Euclidean neighbors; if this graph fails on thin or complex geometry, the method degrades.
  • domain assumption The PU-Net dataset and evaluation protocol provide a valid measure of point cloud super-resolution quality.
    The paper trains and tests on 40 plus 20 Visionair models from PU-Net and uses CD, EMD, F-score, NUC, and Deviation as ground truth. The validity of these proxies is assumed.
  • domain assumption Adversarial training with LS-GAN converges to useful local realism without destabilizing coordinate regression.
    The graph adversarial loss is central to the method, and its stability and benefit are shown empirically rather than proven.
  • domain assumption One-sided Chamfer distance is a valid training objective for point cloud super-resolution.
    The paper removes the symmetric term and uses only Lcd, but its stated justification in Section 3.2 is questionable, so the soundness of this loss choice is an assumption.
  • domain assumption Uniformly sampled ground-truth points from meshes represent ideal high-resolution point clouds.
    The ground truth for training and testing is generated by uniform sampling of mesh surfaces, which is an assumption about what a good HR point cloud should be.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Point Cloud Super Resolution with Adversarial Residual Graph Networks." pith.science (2026). https://pith.science/paper/M23ZTONW

@misc{pith2026190802111,
  author       = {Pith},
  title        = {Pith review of: Point Cloud Super Resolution with Adversarial Residual Graph Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/M23ZTONW}},
  note         = {Machine review of arXiv:1908.02111}
}
read the original abstract

Point cloud super-resolution is a fundamental problem for 3D reconstruction and 3D data understanding. It takes a low-resolution (LR) point cloud as input and generates a high-resolution (HR) point cloud with rich details. In this paper, we present a data-driven method for point cloud super-resolution based on graph networks and adversarial losses. The key idea of the proposed network is to exploit the local similarity of point cloud and the analogy between LR input and HR output. For the former, we design a deep network with graph convolution. For the latter, we propose to add residual connections into graph convolution and introduce a skip connection between input and output. The proposed network is trained with a novel loss function, which combines Chamfer Distance (CD) and graph adversarial loss. Such a loss function captures the characteristics of HR point cloud automatically without manual design. We conduct a series of experiments to evaluate our method and validate the superiority over other methods. Results show that the proposed method achieves the state-of-the-art performance and have a good generalization ability to unseen data.

Figures

Figures reproduced from arXiv: 1908.02111 by the authors.

Figure 1
Figure 1. Point Cloud Super Resolution. (a) is the input LR point cloud with sparse distribution. (b) is the corre￾sponding HR point cloud with dense distribution. (c) and (d) are the HR point cloud generated by PU-Net [37] and our method respectively. Ours is sharper at edges with fewer noisy points. Best viewed in color. data, which have strong assumptions about the underlying surface of the HR point cloud [12, 13]. As for … view at source ↗
Figure 2
Figure 2. Framework Overview. The proposed AR-GCN consists of a generator G and a discriminator D. G is a residual GCN that upsamples the input point cloud progressively with the upsampling ratio 2×. D is also a residual GCN, which learns to distinguish fake HR point cloud from the real one. 3.1. Point Cloud Super Resolution Formally, given a point cloud x with shape n × 3, the goal of point cloud super-resolution is to gener… view at source ↗
Figure 3
Figure 3. Residual Graph Convolution Block. xin is used for querying the k nearest neighbors. xout is the same as xin. The core of graph convolution, G-conv, is defined on a graph G=(υ, ε) and calculated as follows, f p l+1 = w0f p l + w1Σq∈N(p)f q l , ∀p ∈ υ, (6) where w is the learn-able parameters and f p l represents the feature of vertex p at layer l. N(p) is the vertices that con￾nect to p as defined by the adjacency ma… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Qualitative Results. (a) is the LR input point cloud with a sparse distribution, while (b) is the corresponding HR point cloud with a dense distribution. (c) and (d) are the HR point cloud generated by PU-Net [37] and our method. Our results are sharper at edges, which…
Figure 7
Figure 7. Figure 7: Employ our method on real-scanned and un-even [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 6
Figure 6. Figure 6: Mesh Reconstruction from Point Cloud. The differences inside the red boxes are most obvious. Best viewed in color. 3D Reconstruction In 3D reconstruction, the quality and density of the point cloud have a huge impact on the qual￾ity of the reconstructed mesh. However, …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 33 canonical work pages

  1. [1]

    Abadi, P

    M. Abadi, P. Barham, J. Chen, Z. Chen, A. Davis, J. Dean, M. Devin, S. Ghemawat, G. Irving, M. Isard, et al. Tensor- flow: a system for large-scale machine learning. In OSDI,

    work page

  2. [2]

    Alexa, J

    M. Alexa, J. Behr, D. Cohen-Or, S. Fleishman, D. Levin, and C. T. Silva. Computing and rendering point set surfaces. IEEE Transactions on visualization and computer graphics,

    work page

  3. [3]

    Bernardini, J

    F. Bernardini, J. Mittleman, H. Rushmeier, C. Silva, and G. Taubin. The ball-pivoting algorithm for surface recon- struction. IEEE transactions on visualization and computer graphics, 1999. 1, 8

    work page 1999

  4. [4]

    M. M. Bronstein, J. Bruna, Y . LeCun, A. Szlam, and P. Van- dergheynst. Geometric deep learning: going beyond eu- clidean data. IEEE Signal Processing Magazine, 2017. 3

    work page 2017

  5. [5]

    Defferrard, X

    M. Defferrard, X. Bresson, and P. Vandergheynst. Convolu- tional neural networks on graphs with fast localized spectral filtering. In NIPS, 2016. 2

    work page 2016

  6. [6]

    C. Dong, C. C. Loy, K. He, and X. Tang. Learning a deep convolutional network for image super-resolution. InECCV,

    work page

  7. [7]

    H. Fan, H. Su, and L. J. Guibas. A point set generation net- work for 3d object reconstruction from a single image. In CVPR, 2017. 1, 2

    work page 2017

  8. [8]

    Goodfellow, J

    I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio. Gen- erative adversarial nets. In NIPS, 2014. 3

    work page 2014

Show all 39 references
  1. [9]

    W. Han, S. Chang, D. Liu, M. Yu, M. Witbrock, and T. S. Huang. Image super-resolution via dual-state recurrent net- works. In CVPR, 2018. 2

    2018

  2. [10]

    Haris, G

    M. Haris, G. Shakhnarovich, and N. Ukita. Deep backpro- jection networks for super-resolution. In CVPR, 2018. 2

    2018

  3. [11]

    Hartley and A

    R. Hartley and A. Zisserman. Multiple view geometry in computer vision. Cambridge university press, 2003. 1

    2003

  4. [12]

    Huang, D

    H. Huang, D. Li, H. Zhang, U. Ascher, and D. Cohen-Or. Consolidation of unorganized point clouds for surface recon- struction. TOG, 2009. 1

    2009

  5. [13]

    Huang, S

    H. Huang, S. Wu, M. Gong, D. Cohen-Or, U. Ascher, and H. R. Zhang. Edge-aware point set resampling. TOG, 2013. 1, 2

    2013

  6. [14]

    Jiang, Y

    M. Jiang, Y . Wu, and C. Lu. Pointsift: A sift-like net- work module for 3d point cloud semantic segmentation. arXiv:1807.00652, 2018. 2

    2018 arXiv

  7. [15]

    H. Kato, Y . Ushiku, and T. Harada. Neural 3d mesh renderer. In CVPR, 2018. 2

    2018

  8. [16]

    J. Kim, J. Kwon Lee, and K. Mu Lee. Deeply-recursive convolutional network for image super-resolution. In CVPR,

  9. [17]

    D. P. Kingma and J. Ba. Adam: A method for stochastic optimization. arXiv:1412.6980, 2014. 5

    2014 arXiv

  10. [18]

    Ledig, L

    C. Ledig, L. Theis, F. Husz ´ar, J. Caballero, A. Cunningham, A. Acosta, A. P. Aitken, A. Tejani, J. Totz, Z. Wang, et al. Photo-realistic single image super-resolution using a genera- tive adversarial network. In CVPR, 2017. 2

    2017

  11. [19]

    J. Li, B. M. Chen, and G. H. Lee. So-net: Self-organizing network for point cloud analysis. In CVPR, 2018. 2

    2018

  12. [20]

    Y . Li, R. Bu, M. Sun, and B. Chen. Pointcnn. arXiv:1801.07791, 2018. 2

    2018 arXiv

  13. [21]

    Z. Lian, J. Zhang, S. Choi, H. ElNaghy, J. El-Sana, T. Fu- ruya, A. Giachetti, R. A. Guler, L. Lai, C. Li, H. Li, F. A. Limberger, R. Martin, R. U. Nakanishi, A. P. Neto, L. G. Nonato, R. Ohbuchi, K. Pevzner, D. Pickup, P. Rosin, A. Sharf, L. Sun, X. Sun, S. Tari, G. Unal, and...

    2015

  14. [22]

    Lipman, D

    Y . Lipman, D. Cohen-Or, D. Levin, and H. Tal-Ezer. Parameterization-free projection for geometry reconstruc- tion. TOG, 2007. 2

    2007

  15. [23]

    X. Mao, Q. Li, H. Xie, R. Y . Lau, Z. Wang, and S. P. Smol- ley. Least squares generative adversarial networks. In ICCV,

  16. [24]

    Maturana and S

    D. Maturana and S. Scherer. V oxnet: A 3d convolutional neural network for real-time object recognition. In IROS,

  17. [25]

    Pan and etc

    S. Pan and etc. Adversarially regularized graph autoencoder for graph embedding. In IJCAI, 2018. 2

    2018

  18. [26]

    C. R. Qi, H. Su, K. Mo, and L. J. Guibas. Pointnet: Deep learning on point sets for 3d classification and segmentation. arXiv:1612.00593, 2016. 2

    2016 arXiv

  19. [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 NIPS, 2017. 2, 8

    2017

  20. [28]

    Riegler, A

    G. Riegler, A. O. Ulusoy, and A. Geiger. Octnet: Learning deep 3d representations at high resolutions. In CVPR, 2017. 2

    2017

  21. [29]

    Shrivastava, T

    A. Shrivastava, T. Pfister, O. Tuzel, J. Susskind, W. Wang, and R. Webb. Learning from simulated and unsupervised images through adversarial training. In CVPR, 2017. 4

    2017

  22. [30]

    Smith, S

    E. Smith, S. Fujimoto, and D. Meger. 3d object super- resolution. arXiv:1802.09987, 2018. 2

    2018 arXiv

  23. [31]

    Sokolova, N

    M. Sokolova, N. Japkowicz, and S. Szpakowicz. Beyond accuracy, f-score and roc: a family of discriminant measures for performance evaluation. In Australasian joint conference on artificial intelligence, 2006. 5

    2006

  24. [32]

    H. Su, V . Jampani, D. Sun, S. Maji, E. Kalogerakis, M.-H. Yang, and J. Kautz. Splatnet: Sparse lattice networks for point cloud processing. In CVPR, 2018. 2

    2018

  25. [33]

    X. Sun, J. Wu, X. Zhang, Z. Zhang, C. Zhang, T. Xue, J. B. Tenenbaum, and W. T. Freeman. Pix3d: Dataset and methods for single-image 3d shape modeling. In CVPR, 2018. 2

    2018

  26. [34]

    N. Wang, Y . Zhang, Z. Li, Y . Fu, W. Liu, and Y .-G. Jiang. Pixel2mesh: Generating 3d mesh models from single rgb im- ages. arXiv:1804.01654, 2018. 2

    2018 arXiv

  27. [35]

    Y . Wang, L. Wang, H. Wang, and P. Li. End-to-end im- age super-resolution via deep and shallow convolutional net- works. arXiv:1607.07680, 2016. 2

    2016 arXiv

  28. [36]

    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, 2015. 2, 8

    2015

  29. [37]

    L. Yu, X. Li, C.-W. Fu, D. Cohen-Or, and P.-A. Heng. Pu- net: Point cloud upsampling network. In CVPR, 2018. 1, 2, 5, 6, 7, 8

    2018

  30. [38]

    Zhang and M

    Y . Zhang and M. Rabbat. A graph-cnn for 3d point cloud classification. In ICASSP, 2018. 2

    2018

  31. [39]

    Zhang, Y

    Y . Zhang, Y . Tian, Y . Kong, B. Zhong, and Y . Fu. Residual dense network for image super-resolution. In CVPR, 2018. 2

    2018

Pith tools

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