Deep Meta Functionals for Shape Representation

Gidi Littwin; Lior Wolf

REVIEW 3 major objections 4 minor 37 references

Deep Meta Functionals for Shape Representation

T0 review · 3 major / 4 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 This paper reports that a hypernetwork which maps a single image to the weights of a point-classifying MLP reconstructs 3D shapes more accurately than voxel-, silhouette-, and mesh-based baselines, while providing unlimited output…

desk verdict The paper's core idea—turning an image directly into the weights of a per-shape occupancy MLP—is genuinely novel and well tested, but its headline accuracy claim rests on borrowed baseline numbers and its smooth-manifold guarantee is overstated. read the letter →

arxiv 1908.06277 v1 pith:WJ7DFWTZ submitted 2019-08-17 cs.CV cs.LG

Gidi Littwin , Lior Wolf This is my paper · ORCID

classification cs.CVcs.LG

keywords single-view3DreconstructionimplicitsurfaceshypernetworkslevelsetsoccupancyclassificationshaperepresentationNetdifferentiablerendering

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

Deep Meta Functionals recasts single-view 3D reconstruction as two nested networks: an image encoder $f$ produces the weights of a small MLP $g$, and $g$ classifies every point in space as inside or outside the object. The reconstructed shape is the level set $g(p)=0.5$, so the representation inherits continuous gradients, arbitrary resolution, and free topology. The paper shows that this single-loss scheme outperforms existing voxel-, silhouette-, and mesh-based methods on two ShapeNet benchmarks, at $32^3$ and $256^3$ grid resolutions, and also gives a lower Chamfer distance than AtlasNet. If correct, the method establishes weight-prediction as a viable representation family for shape inference, not just a training trick.

What carries the argument

The machinery is a hypernetwork $f$ that outputs all parameters of a per-shape occupancy network $g$, including per-layer scale vectors $\theta_s^{(n)}$ that multiply the pre-activation output of each layer. Because $g$ is an ELU-activated MLP, it is differentiable, and the paper invokes the implicit function theorem to claim the $0.5$ level set is a smooth manifold. The capacity argument transplants the linear-regions bound for ReLU networks (see [30]) to $g$: for the tested architectures the number of linear regions lies between $10^4$ and $8.6\times10^{19}$, which is the concrete sense in which the shape representation has high capacity — the field can be sampled at any density and polygonized with marching cubes [27].

What would settle it

Take the released model, scan the $0.5$ level set of $g$ for points where the gradient with respect to $p$ vanishes or where the field is locally constant across the threshold, and count how often this occurs over the ShapeNet test set. If degenerate points appear in a nontrivial fraction of shapes, the smooth-manifold guarantee is false as stated, even if the IoU numbers still stand.

Watch

Extended reading notes

Core claim

The paper's central claim is that a shape can be represented as the decision boundary of a classifier whose weights are generated on the fly from an image. Concretely, $f(I,\theta_f)=\theta_I$ and $s_p^I=g(p,\theta_I)$, where $g$ is a four-hidden-layer MLP with about 3394 parameters and $f$ is a ResNet-style CNN with multiple linear heads, one per layer of $g$. Training uses a single cross-entropy loss evaluated on sampled points, with extra Gaussian samples near ground-truth mesh vertices to concentrate effort near the boundary. The reported results are mean IoU 69.1 versus 64.8 for the best prior method at $32^3$ resolution, 66.2 versus 65.5 at $256^3$ resolution, and average Chamfer distance $4.35\times10^{-3}$ versus $9.52\times10^{-3}$ for AtlasNet on the same protocol. The paper also shows that averaging the embedding over multiple test views improves accuracy, and that linearly interpolating embeddings produces smooth cross-class shape blends.

Load-bearing premise

The load-bearing premise is that $0.5$ is a regular value of the trained point classifier, which the paper asserts from differentiability but never verifies; if the field flattens at the threshold, the extracted surface need not be a smooth manifold.

Editorial extensions

If this is right

A single model trained without class labels beats prior baselines on 12 of 13 categories at $32^3$ resolution and on 8 of 13 at $256^3$, so the representation itself, not a class prior, drives the accuracy.
Because the output is a continuous scalar field, the same trained network can be rendered at any grid density; the paper renders at $128^3$ for display, but the field can be sampled at finer or coarser resolution at test time.
Boundary-focused sampling improves mean IoU from 63.5 to 65.8 over uniform sampling, confirming that the differentiable field benefits from gradient information concentrated near the surface.
Averaging the embedding over multiple views at test time improves IoU monotonically and outperforms the multi-view-trained 3D-R2N2 baseline without any retraining for the multi-view setting.
The representation extends to dynamic shapes by adding a time coordinate to $g$'s input, with a proof-of-concept showing a single digit image generating an entire temporal sequence.

Reading between the lines

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

Beyond the paper, the smooth-manifold guarantee should be read as conditional: it holds only where $0.5$ is a regular value, and the paper never checks that, so a more cautious claim is that the surface is a manifold except at degenerate threshold points.
One testable extension, not in the paper, is to count distinct connected components or handle holes in the predicted surface across ShapeNet classes; if the topology is genuinely arbitrary, the method should reconstruct objects with holes (e.g., mugs, stools) without template restrictions.
Another extension is to compare weight-prediction against embedding-conditioned implicit decoders (e.g., DeepSDF or Occupancy Networks) with matched decoder sizes and training budgets; the paper argues weight prediction is more parameter-efficient, but does not run that controlled comparison.
A practical check of the resolution claim: train on low-resolution voxel supervision, then render at $512^3$ and measure whether IoU against a high-resolution ground truth stays flat; if it degrades, the 'unlimited resolution' claim is bounded by what the hypernetwork learns.

Share X Bluesky LinkedIn Reddit HN

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

You should know this paper for its mechanism: instead of conditioning a decoder on an embedding (like Occupancy Networks or DeepSDF), it predicts the decoder's weights directly from the image. That is a real twist, and the paper shows it works about as well as one would hope. The method is simple, the loss is a single cross-entropy, and the code is released.

What's good: The experiments are reasonably thorough for 2019: two benchmarks, comparisons against voxel, silhouette, point cloud, and mesh baselines, plus capacity and boundary-sampling ablations. The qualitative results look plausible, and the multi-view embedding averaging is a nice touch. The capacity argument via linear regions is borrowed but correctly applied.

The soft spots are exactly where the stress-test note lands. Tables 2-4 use numbers from prior papers, not re-computed under a common harness. There are no error bars or multiple seeds. The most relevant concurrent methods—Occupancy Networks and DeepSDF—are discussed in Section 2 but never appear in the quantitative tables. The 256^3 result over VP3D is 66.2 vs 65.5 mean IoU: a 0.7-point lead that is well within typical evaluation-protocol spread. So the abstract's 'more accurate than the existing methods' is not as solid as the prose implies. That is a load-bearing claim, and the margin is thin enough that re-running baselines could flip it.

Second, Section 4's smooth-manifold guarantee is stated too strongly. The implicit function theorem gives a smooth manifold only if 0.5 is a regular value of the trained g. The paper never checks that, and the Jacobian-norm visualization in Sec 5.6 actually hints at near-flat regions. This is a real gap but secondary to the empirical claim—even a non-manifold level set could score fine on IoU.

One thing to push back on: the stress-test note's worry about the regularity gap being secondary is right. And the reader's circularity burden of 0.0 is appropriate—there is no equation in the paper fitted to the reported results.

Who it's for: anyone working on implicit shape representations or hypernetworks. It is a useful data point and a distinct mechanism that deserves citation. It is not a breakthrough that reshapes the field, but it is a solid, honest piece of work.

Recommendation: a serious editor should send this to peer review. The right outcome is likely major revision: either re-run the key baselines under one evaluation harness or remove the 'more accurate than existing methods' claim and present results conditionally. The method itself is valuable regardless of the precise ranking.

Referee Report

3 major / 4 minor

Summary. The paper introduces a hypernetwork-based representation for single-view 3D reconstruction. A convolutional network f maps an input image to the weights of a small MLP g; the reconstructed shape is the 0.5 level set of g, so each shape is represented by a per-instance implicit classifier. Training uses a single cross-entropy loss with importance sampling near the ground-truth mesh surface, and rendering is performed by evaluating g on a grid followed by marching cubes. Experiments on ShapeNet-core compare the method with voxel-, point-cloud-, and mesh-based baselines at 32^3 and 256^3 resolutions, reporting improvements in mean IoU in most categories, a large improvement in Chamfer distance over two baselines, ablations over the architectures of f and g, a comparison of boundary sampling with uniform sampling, multi-view embedding averaging, and qualitative results including linear interpolations between shape embeddings.

Significance. If the empirical claims hold, this is a conceptually clean and influential idea: rather than conditioning a fixed implicit decoder on a latent code, the image directly produces the decoder weights, yielding a per-shape network with effectively unlimited output resolution and no template-topology restriction. The paper explicitly ships code, provides a substantial set of ablations (activation, width, depth, encoder size, sampling scheme), and demonstrates attractive qualitative properties such as multi-view fusion and smooth interpolation. The central limitation is that the main quantitative conclusion is built on comparisons with baseline numbers taken from prior papers without a shared evaluation harness or variance estimates, and the most closely related concurrent implicit-function baselines are discussed but not evaluated. The theoretical smoothness claim also depends on an unverified regularity condition. These issues are fixable within the scope of the paper, so the work is potentially publishable after a major revision that re-grounds the empirical comparison and sharpens the theoretical statements.

major comments (3)

[Section 5.2, Tables 2 and 3] The central superiority claim is supported only by point estimates taken from prior publications for the baselines, with no reported per-seed variance and no unified evaluation harness. In Table 3 the mean-IoU lead over VP3D is 66.2 versus 65.5, which is smaller than typical seed-to-seed variation for this type of reconstruction, so the claimed margin may not survive re-running the baselines with identical preprocessing and an identical IoU script. Please re-evaluate the baselines under the same protocol, report multiple seeds, and provide error bars or confidence intervals for all quantitative comparisons.
[Section 2 and Tables 2-4] The paper discusses Occupancy Networks [29], DeepSDF [31], and IM-Net [7] as concurrent implicit-function methods, and Section 2 criticizes their training cost and generalization, but none of these methods appears in the quantitative tables even though [29] reports multi-class ShapeNet results. Because the abstract claims superiority over existing methods, the quantitative comparison must include at least one of these directly related baselines under the same split and evaluation protocol; otherwise the reader cannot distinguish representation gains from dataset or evaluation differences.
[Section 4] The smooth-manifold guarantee invoked from the implicit function theorem requires 0.5 to be a regular value of the trained network g; this condition is assumed but never checked, and a trained classifier can have flat regions or degenerate critical points at the threshold. Please either verify regularity empirically (e.g., by computing the Jacobian norm on the extracted surface and checking that it does not vanish) or state the claim as conditional. In addition, the capacity argument uses an upper bound on the number of linear regions, not a lower bound on representable decision boundaries, so the phrase that the surface capacity is exponential in the network size overstates what the cited bound establishes.

minor comments (4)

[Section 3, Eq. (4)] The indexing of the generated parameters, e.g., θW(n)_I, θs(n)_I, and θb(n)_I, is hard to parse; please define the exact shapes of all tensors and make the layer indexing consistent throughout.
[Section 5.1 and Section 5.6] There are several typos: 'H¨anee et al' should be 'H¨ane et al', 'Ths obtained norm' should be 'The obtained norm', and Reference [4] is duplicated with an inconsistent author spelling.
[Section 3, Architecture] The text states that the default g has 3394 parameters, but the configuration is described as four hidden layers of 32 units; please show the parameter count computation or specify which layer types (weights, bias, scale) are included, as the count is not obvious from the description.
[Section 6, Differentiable rendering proof of concept] The silhouette-based proof-of-concept reports IoU 64.4 versus 60.0 for [20] but immediately notes that the three views are different; this comparison should either be removed or clearly labeled as non-comparable, since it currently reads as a quantitative advantage.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the empirical claim is an independent benchmark comparison and the theoretical statements are quoted standard results, not definitions of the reported outcomes.

full rationale

The paper's central claim is empirical: a hypernetwork that predicts the weights of a per-shape occupancy MLP outperforms prior voxel-, silhouette-, and mesh-based methods on ShapeNet single-view reconstruction. This claim is supported by quantitative comparisons in Tables 2-4 against externally published baselines on standard train/test splits (Choy et al. and Hane et al. datasets). No equation in the paper is fitted to those reported IoU or Chamfer numbers, and the reported method is trained with a standard cross-entropy loss (Eq. 3) against ground-truth occupancy. The capacity statement is quoted from Montufar et al. and is not used to define the accuracy results. The smooth-manifold claim in Section 4 relies on the implicit function theorem, conditional on the unverified regular-value assumption; this is a correctness gap, not a circular reduction. The only self-citation (Ref. 23, a dynamic convolutional layer) is historical background for hypernetworks and does not carry any load-bearing argument. Therefore no prediction or first-principles result reduces by construction to its own inputs.

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

No new physical entities are postulated. The central empirical claim rests on standard deep-learning axioms plus specific assumptions about the occupancy field and the regular-value condition; see individual entries.

free parameters (4)

g hidden layers L and width = L=4, 32 hidden units (default)
Chosen by hand; sensitivity study in Table 5 shows performance is flat across 3-6 layers and 16-64 units, so the exact values are not load-bearing.
boundary sampling variance = 0.1
Gaussian variance around ground truth vertices; chosen by hand, not fitted to the target result.
weight of uniform samples = 10%
Fraction of uniformly sampled points added per batch; chosen by hand.
image encoder architecture = N=64, B=5, K=2
Chosen by hand; Table 6 shows insensitivity across N and B, so the exact values are not load-bearing.

assumptions (4)

domain assumption Level set at threshold 0.5 is a regular value for the trained classifier
Section 4 invokes the implicit function theorem to guarantee a smooth manifold, which requires nonzero gradient at the 0.5 surface; this is not verified for trained networks.
domain assumption Ground-truth occupancy labels computed from 32^3 voxel grids are accurate enough to supervise the continuous field
Used in Section 3 for label computation; coarse voxelization may provide biased supervision near thin structures.
domain assumption Montufar et al. bound on linear regions is a valid proxy for shape-representation capacity
Section 4 equates the number of linear regions of a ReLU network with the capacity to represent shape boundaries, which conflates classification power with geometric fidelity.
domain assumption ShapeNet meshes are closed, watertight manifolds
Needed so inside/outside labels are well-defined; not true for all CAD models.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Deep Meta Functionals for Shape Representation." pith.science (2026). https://pith.science/paper/WJ7DFWTZ

@misc{pith2026190806277,
  author       = {Pith},
  title        = {Pith review of: Deep Meta Functionals for Shape Representation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WJ7DFWTZ}},
  note         = {Machine review of arXiv:1908.06277}
}

read the original abstract

We present a new method for 3D shape reconstruction from a single image, in which a deep neural network directly maps an image to a vector of network weights. The network \textcolor{black}{parametrized by} these weights represents a 3D shape by classifying every point in the volume as either within or outside the shape. The new representation has virtually unlimited capacity and resolution, and can have an arbitrary topology. Our experiments show that it leads to more accurate shape inference from a 2D projection than the existing methods, including voxel-, silhouette-, and mesh-based methods. The code is available at: https://github.com/gidilittwin/Deep-Meta

Figures

Figures reproduced from arXiv: 1908.06277 by the authors.

**Figure 2.** The architecture of our neural networks. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** A t-SNE visualizations of object embedding from the 13 [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (6 more)

**Figure 4.** Figure 4: Linear shape interpolation between objects of the same [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

**Figure 6.** Figure 6: Shape surface extracted with different thresholds on [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]

**Figure 8.** Figure 8: Learning curves for ShapeNet showing mean IOU vs. [PITH_FULL_IMAGE:figures/full_fig_p006_8.png]

**Figure 9.** Figure 9: Reconstruction from real-world images. (left) input im [PITH_FULL_IMAGE:figures/full_fig_p007_9.png]

**Figure 12.** Figure 12: A sequence of temporal reconstruction results obtained [PITH_FULL_IMAGE:figures/full_fig_p008_12.png]

**Figure 11.** Figure 11: Single Image 3D Reconstruction. (rows 1,3) input im [PITH_FULL_IMAGE:figures/full_fig_p008_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 30 canonical work pages

[29]

Occupancy networks: Learning 3d reconstruction in function space, 2018

Lars Mescheder, Michael Oechsle, Michael Niemeyer, Se- bastian Nowozin, and Andreas Geiger. Occupancy networks: Learning 3d reconstruction in function space, 2018. 2

work page 2018
[31]

Deepsdf: Learning con- tinuous signed distance functions for shape representation,

Jeong Joon Park, Peter Florence, Julian Straub, Richard Newcombe, and Steven Lovegrove. Deepsdf: Learning con- tinuous signed distance functions for shape representation,
[7]

Learning implicit ﬁelds for generative shape modeling, 2018

Zhiqin Chen and Hao Zhang. Learning implicit ﬁelds for generative shape modeling, 2018. 2

work page 2018
[1]

Learning feed-forward one-shot learners

Luca Bertinetto, Jo ˜ao F Henriques, Jack Valmadre, Philip Torr, and Andrea Vedaldi. Learning feed-forward one-shot learners. In Advances in Neural Information Processing Sys- tems, pages 523–531, 2016. 2

work page 2016
[2]

Polygonization of implicit surfaces

Jules Bloomenthal. Polygonization of implicit surfaces. Computer Aided Geometric Design , 5(4):341–355, 1988. 8

work page 1988
[3]

A geometric model for active contours in image pro- cessing

Vicent Caselles, Francine Catt ´e, Tomeu Coll, and Franc ¸oise Dibos. A geometric model for active contours in image pro- cessing. Numerische mathematik, 66(1):1–31, 1993. 2

work page 1993
[4]

A geometric model for active contours in image pro- cessing

Vicent Caselles, Francine Catt, Bartomeu Coll, and Franoise Dibos. A geometric model for active contours in image pro- cessing. Numerische Mathematik, 66:1–31, 01 1993. 8

work page 1993
[5]

Geodesic active contours

Vicent Caselles, Ron Kimmel, and Guillermo Sapiro. Geodesic active contours. International journal of computer vision, 22(1):61–79, 1997. 2

work page 1997

Show all 37 references

[6]

Chang, Thomas Funkhouser, Leonidas Guibas, Pat Hanrahan, Qixing Huang, Zimo Li, Silvio Savarese, Mano- lis Savva, Shuran Song, Hao Su, Jianxiong Xiao, Li Yi, and Fisher Yu

Angel X. Chang, Thomas Funkhouser, Leonidas Guibas, Pat Hanrahan, Qixing Huang, Zimo Li, Silvio Savarese, Mano- lis Savva, Shuran Song, Hao Su, Jianxiong Xiao, Li Yi, and Fisher Yu. Shapenet: An information-rich 3d model reposi- tory, 2015. 1

2015
[8]

3d-r2n2: A uniﬁed approach for single and multi-view 3d object reconstruction

Christopher B Choy, Danfei Xu, JunYoung Gwak, Kevin Chen, and Silvio Savarese. 3d-r2n2: A uniﬁed approach for single and multi-view 3d object reconstruction. In European conference on computer vision , pages 628–644. Springer,
[9]

Fast and accurate deep network learning by exponential linear units (elus)

Djork-Arn ´e Clevert, Thomas Unterthiner, and Sepp Hochre- iter. Fast and accurate deep network learning by exponential linear units (elus). arXiv preprint arXiv:1511.07289 , 2015. 4

2015 arXiv
[10]

Haoqiang Fan, Hao Su, and Leonidas J. Guibas. A point set generation network for 3d object reconstruction from a single image. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), July 2017. 1, 7

2017
[11]

Kim, Bryan C

Thibault Groueix, Matthew Fisher, Vladimir G. Kim, Bryan C. Russell, and Mathieu Aubry. A papier-mache ap- proach to learning 3d surface generation. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition , Jun 2018. 2, 6, 7

2018
[12]

Hypernetworks

David Ha, Andrew Dai, and Quoc V Le. Hypernetworks. arXiv preprint arXiv:1609.09106, 2016. 2

2016 arXiv
[13]

Hi- erarchical surface prediction for 3d object reconstruction

Christian H ¨ane, Shubham Tulsiani, and Jitendra Malik. Hi- erarchical surface prediction for 3d object reconstruction. In 2017 International Conference on 3D Vision (3DV) , pages 412–420. IEEE, 2017. 1, 5, 6, 7

2017
[14]

Ray tracing algebraic surfaces

Pat Hanrahan. Ray tracing algebraic surfaces. In ACM SIGGRAPH Computer Graphics , volume 17, pages 83–90. ACM, 1983. 8

1983
[15]

Delving deep into rectiﬁers: Surpassing human-level perfor- mance on imagenet classiﬁcation

Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Delving deep into rectiﬁers: Surpassing human-level perfor- mance on imagenet classiﬁcation. InProceedings of the 2015 IEEE International Conference on Computer Vision (ICCV), ICCV ’15, pages 1026–1034, Washington, DC, USA...

2015
[16]

Dynamic ﬁlter networks

Xu Jia, Bert De Brabandere, Tinne Tuytelaars, and Luc V Gool. Dynamic ﬁlter networks. In Advances in Neural In- formation Processing Systems, pages 667–675, 2016. 2

2016
[17]

Gal: Geometric adversarial loss for single-view 3d-object recon- struction

Li Jiang, Shaoshuai Shi, Xiaojuan Qi, and Jiaya Jia. Gal: Geometric adversarial loss for single-view 3d-object recon- struction. In The European Conference on Computer Vision (ECCV), September 2018. 1

2018
[18]

Learning a multi-view stereo machine, 2017

Abhishek Kar, Christian Hne, and Jitendra Malik. Learning a multi-view stereo machine, 2017. 6, 7

2017
[19]

Snakes: Active contour models

Michael Kass, Andrew Witkin, and Demetri Terzopoulos. Snakes: Active contour models. International journal of computer vision, 1(4):321–331, 1988. 2

1988
[20]

Learning view pri- ors for single-view 3d reconstruction

Hiroharu Kato and Tatsuya Harada. Learning view pri- ors for single-view 3d reconstruction. arXiv preprint arXiv:1811.10719, 2018. 6, 7, 8

2018 arXiv
[21]

Neu- ral 3d mesh renderer

Hiroharu Kato, Yoshitaka Ushiku, and Tatsuya Harada. Neu- ral 3d mesh renderer. InProceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 3907– 3916, 2018. 1

2018
[22]

Gradient ﬂows and ge- ometric active contour models

Satyanad Kichenassamy, Arun Kumar, Peter Olver, Allen Tannenbaum, and Anthony Yezzi. Gradient ﬂows and ge- ometric active contour models. In Proceedings of IEEE In- ternational Conference on Computer Vision, pages 810–815. IEEE, 1995. 2

1995
[23]

A dynamic convolutional layer for short range weather prediction

Benjamin Klein, Lior Wolf, and Yehuda Afek. A dynamic convolutional layer for short range weather prediction. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 4840–4848, 2015. 2

2015
[24]

Kosinski

A.A. Kosinski. Differential Manifolds. Dover Book on Math- ematics. Dover Publications, 2007. 4

2007
[25]

Active contour based segmentation of 3d surfaces

Matthias Krueger, Patrice Delmas, and Georgy Gimelfarb. Active contour based segmentation of 3d surfaces. In Eu- ropean Conference on Computer Vision , pages 350–363. Springer, 2008. 2

2008
[26]

Soft rasterizer: Differentiable rendering for unsupervised single- view mesh reconstruction

Shichen Liu, Weikai Chen, Tianye Li, and Hao Li. Soft rasterizer: Differentiable rendering for unsupervised single- view mesh reconstruction. arXiv preprint arXiv:1901.05567,

1901 arXiv
[27]

Lorensen and Harvey E

William E. Lorensen and Harvey E. Cline. Marching cubes: A high resolution 3d surface construction algorithm. SIG- GRAPH Comput. Graph., 21(4):163–169, Aug. 1987. 4

1987
[28]

Evolu- tionary fronts for topology-independent shape modeling and recovery

Ravi Malladi, James A Sethian, and Baba C Vemuri. Evolu- tionary fronts for topology-independent shape modeling and recovery. In European conference on Computer vision, pages 1–13. Springer, 1994. 2

1994
[30]

On the number of linear regions of deep neural networks

Guido F Montufar, Razvan Pascanu, Kyunghyun Cho, and Yoshua Bengio. On the number of linear regions of deep neural networks. In Advances in neural information process- ing systems, pages 2924–2932, 2014. 4

2014
[32]

Matryoshka networks: Predicting 3d geometry via nested shape layers

Stephan R Richter and Stefan Roth. Matryoshka networks: Predicting 3d geometry via nested shape layers. In Proceed- ings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 1936–1944, 2018. 1, 7

1936
[33]

Riegler, S

G. Riegler, S. Schulter, M. Rther, and H. Bischof. Condi- tioned regression models for non-blind single image super- resolution. In 2015 IEEE International Conference on Com- puter Vision (ICCV), pages 522–530, Dec 2015. 2

2015
[34]

Octnet: Learning deep 3d representations at high resolu- tions

Gernot Riegler, Ali Osman Ulusoy, and Andreas Geiger. Octnet: Learning deep 3d representations at high resolu- tions. 2017 IEEE Conference on Computer Vision and Pat- tern Recognition (CVPR), Jul 2017. 1

2017
[35]

Octree generating networks: Efﬁcient convolutional archi- tectures for high-resolution 3d outputs

Maxim Tatarchenko, Alexey Dosovitskiy, and Thomas Brox. Octree generating networks: Efﬁcient convolutional archi- tectures for high-resolution 3d outputs. InProceedings of the IEEE International Conference on Computer Vision , pages 2088–2096, 2017. 1, 7

2017
[36]

Pixel2mesh: Generating 3d mesh models from single rgb images

Nanyang Wang, Yinda Zhang, Zhuwen Li, Yanwei Fu, Wei Liu, and Yu-Gang Jiang. Pixel2mesh: Generating 3d mesh models from single rgb images. In Proceedings of the Euro- pean Conference on Computer Vision (ECCV), pages 52–67,
[37]

Inferring point clouds from single monocular images by depth intermedia- tion, 2018

Wei Zeng, Sezer Karaoglu, and Theo Gevers. Inferring point clouds from single monocular images by depth intermedia- tion, 2018. 5, 7

2018

Pith tools

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

[1] [29]

Occupancy networks: Learning 3d reconstruction in function space, 2018

Lars Mescheder, Michael Oechsle, Michael Niemeyer, Se- bastian Nowozin, and Andreas Geiger. Occupancy networks: Learning 3d reconstruction in function space, 2018. 2

work page 2018

[2] [31]

Deepsdf: Learning con- tinuous signed distance functions for shape representation,

Jeong Joon Park, Peter Florence, Julian Straub, Richard Newcombe, and Steven Lovegrove. Deepsdf: Learning con- tinuous signed distance functions for shape representation,

[3] [7]

Learning implicit ﬁelds for generative shape modeling, 2018

Zhiqin Chen and Hao Zhang. Learning implicit ﬁelds for generative shape modeling, 2018. 2

work page 2018

[4] [1]

Learning feed-forward one-shot learners

Luca Bertinetto, Jo ˜ao F Henriques, Jack Valmadre, Philip Torr, and Andrea Vedaldi. Learning feed-forward one-shot learners. In Advances in Neural Information Processing Sys- tems, pages 523–531, 2016. 2

work page 2016

[5] [2]

Polygonization of implicit surfaces

Jules Bloomenthal. Polygonization of implicit surfaces. Computer Aided Geometric Design , 5(4):341–355, 1988. 8

work page 1988

[6] [3]

A geometric model for active contours in image pro- cessing

Vicent Caselles, Francine Catt ´e, Tomeu Coll, and Franc ¸oise Dibos. A geometric model for active contours in image pro- cessing. Numerische mathematik, 66(1):1–31, 1993. 2

work page 1993

[7] [4]

A geometric model for active contours in image pro- cessing

Vicent Caselles, Francine Catt, Bartomeu Coll, and Franoise Dibos. A geometric model for active contours in image pro- cessing. Numerische Mathematik, 66:1–31, 01 1993. 8

work page 1993

[8] [5]

Geodesic active contours

Vicent Caselles, Ron Kimmel, and Guillermo Sapiro. Geodesic active contours. International journal of computer vision, 22(1):61–79, 1997. 2

work page 1997