Pith. sign in

REVIEW 4 major objections 6 minor 35 references

Fixing the Perspective: A Critical Examination of Zero-1-to-3

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Zero-1-to-3's image-conditioning cross-attention does not actually attend: with a single 1×768 context vector supplying keys and values, softmax weights become uniformly 1 and the layer degenerates into copying that vector, so whatever…

desk verdict A plausible but unproven diagnosis of Zero-1-to-3's cross-attention collapse, with no experimental support; reject as-is, but worth a short technical note if the premise is verified. read the letter →

arxiv 2411.15706 v1 pith:WOARI6UM submitted 2024-11-24 cs.CV cs.LG

Jack Yu , Xueying Jia , Charlie Sun , Prince Wang This is my paper · ORCID
classification cs.CVcs.LG
keywords novelviewsynthesisZero-1-to-3cross-attentionlatentdiffusionmodelsimage-to-3Dgenerationattentioncollapseconditioningembeddingsmulti-viewconsistency
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 argues that Zero-1-to-3, a widely used model for turning one image into a 3D object, contains a silent architectural failure in its conditioning pathway. Within the cross-attention layer of its diffusion UNet, the image-and-camera context is a single vector in $\mathbb{R}^{1\times 768}$; because softmax over one key is always 1, the attention mechanism cannot weight any feature and reduces to duplicating that vector. The paper concludes that the model's conditioning signal never actually gets attended to, and that the residual connection carries the information as a plain additive term — an expressive bottleneck inside an otherwise attention-based network. To fix this, the authors propose two changes: a multi-view architecture that processes several conditioning images at once, and a "revamped embedding" that concatenates image and pose features along the token dimension so true cross-attention can operate. Their own experiments are preliminary and, by their admission, not fully trained or metric-evaluated, which matters for judging the fixes but not for the core claim about the released model.

What carries the argument

The load-bearing object is the cross-attention head inside the Spatial Transformer of Zero-1-to-3's conditional diffusion UNet: the hidden state produces queries $Q$, while the single combined image-and-pose context vector $c\in\mathbb{R}^{768}$ produces the keys and values $K,V$. The identity that carries the whole argument is the softmax collapse with one key: since softmax over a single element equals 1, attention degenerates to $\mathrm{Attention}(Q,K,V)=V$, i.e., the context vector gets duplicated to every spatial position. This identity makes the $Q$ and $K$ projections computationally redundant and forces the model to carry conditioning information through the residual connection as an additive term; the paper's two proposed fixes — a token-dimension (vertical) concatenation of image and pose embeddings, and a multi-view context aggregation — are both designed to give the attention mechanism more than one key so the collapse cannot occur.

What would settle it

Print the shape of the context tensor at the cross-attention input of the released Zero-1-to-3 model (or instrument the forward pass at the cited line): a sequence length of 1 confirms the premise; a length greater than 1 refutes it. Independently, load the released weights, feed a fixed hidden state, and vary a single coordinate of the context vector: under the collapse claim the output changes only through the residual path, and perturbing features that enter only through K or Q produces exactly zero change in the output.

Watch

Extended reading notes

Core claim

The central claim is that the cross-attention layer in Zero-1-to-3's UNet Spatial Transformer is computationally degenerate. The context embedding $c=f(\mathrm{CLIP}(I),P)$, where $I$ is the input image and $P=(r,\phi,\theta)$ the camera pose, is a single vector of width 768. When this vector supplies the keys and values while the hidden state supplies the queries, the attention identity becomes $\mathrm{Attention}(Q,K,V)=\mathrm{softmax}(QK^T/\sqrt{d_k})V = V$, because the softmax over a single key is identically 1. The layer therefore duplicates $V$ for every query position, and the $Q$ and $K$ projection layers do no work at all. The paper locates this behavior in the released code at line 189 and argues that the model's observed conditioning ability comes from the Transformer residual connection, which adds the context vector to the hidden state — a strictly weaker operation than attention. On this diagnosis, the paper proposes making cross-attention real by concatenating the image and camera-pose embeddings vertically (along the token dimension) instead of horizontally, plus a multi-view extension that combines several conditioning views; it reports initial results for the revamped embedding but leaves full training and metric evaluation to future work.

Load-bearing premise

Everything rests on the claim that the released Zero-1-to-3 code actually feeds its cross-attention layer a single $1\times 768$ context vector as both keys and values; the paper pins this to one code line that it does not reproduce, so if the real implementation uses several context tokens, or expands or permutes the tensor before attention, the collapse does not occur.

Editorial extensions

If this is right

  • If the collapse claim is right, published Zero-1-to-3 results should be re-described: the conditioning image affects generation through a linear projection plus an additive residual, not through selective attention, so the model never chooses which input features matter for a target view.
  • A corrected embedding that gives cross-attention real tokens removes the conditioning bottleneck and makes the attention pathway strictly more expressive for the same trained weights.
  • The multi-view variant, once trained, should keep gaining information from each additional conditioning view instead of saturating at a single pooled vector — the paper's stated motivation for improving consistency on occluded and back views.
  • The model's known failure modes — inconsistent back views and artifacts in occluded regions — are consistent with a conditioning-capacity bottleneck, so the fix targets the cause rather than adding more data.

Reading between the lines

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

  • Editorial: the same single-context-vector pattern likely appears in other models that pool CLIP embeddings before cross-attention, and a shape check on the context tensor at the attention input would reveal how widespread the degeneracy is.
  • Editorial: because the collapse makes attention position-independent, the model cannot express view-dependent conditioning (say, "the handle is on the left in this view"); restoring a real token dimension may unlock spatially targeted conditioning that the current architecture structurally cannot represent.
  • Editorial: a retraining-free confirmation exists — feed the released weights a fixed hidden state and vary one coordinate of the context vector; if the output changes only through the residual path, the collapse holds.
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.

Referee Report

4 major / 6 minor

Summary. The paper examines the cross-attention mechanism of the Zero-1-to-3 novel-view-synthesis model and claims to uncover a critical discrepancy between its theoretical design and implementation: the image-conditioning context is a single 1x768 vector, so softmax attention over a single key collapses to weight 1, reducing cross-attention to a linear projection of the context. Based on this diagnosis, the authors propose two improvements: a multi-view conditioning architecture and a revised embedding that concatenates image and camera-angle features vertically. Experiments are conducted on a 25% subset of Objaverse for the revised-embedding model only. The paper reports that the generated outputs have underwhelming quality and that quantitative metrics are skipped, and it concludes with limitations and future work.

Significance. The mathematical observation in Section 5.2 is correct: with a single context vector, the softmax weights are identically 1, so the cross-attention output is a learned linear transformation of the context and the Q and K projections do not affect the output. This is a useful diagnostic observation about Zero-1-to-3's design. However, the paper does not provide verifiable evidence that the premise holds in the released implementation, and it offers no quantitative validation of either proposed fix. The claimed improvements are explicitly contradicted by the paper's own Section 7.3. The significance of the work is therefore diagnostic rather than a demonstrated methodological contribution.

major comments (4)
  1. [Abstract and §7.3] The abstract and contributions claim that the proposed changes yield 'potential improvements' and 'effective utilization' of cross-attention, but Section 7.3 states that the proposed models have 'underwhelming generation quality' and that 'we will skip the metrics evaluation.' Table 2, which is supposed to summarize the results, contains no actual result cells. There is thus no quantitative evidence supporting the central claim of improved synthesis, and the paper's own text contradicts it.
  2. [§5.3] The single-context-vector premise, on which the entire collapse derivation and both proposed fixes depend, is supported only by a reference to 'line 189' of the Zero-1-to-3 repository. The manuscript provides no code excerpt, tensor-shape trace, or commit identifier, so a reader cannot verify that the context really has sequence length 1 in the released implementation. Because this premise is load-bearing, the evidence must be made verifiable.
  3. [§6.5 and §6.6] The multi-view architecture, presented as one of the two main contributions, is explicitly not implemented: Section 6.5 says 'we have postponed full implementation of this architecture,' and Section 6.6 lists only 'expected outcomes.' The paper therefore provides no empirical or simulation-based evidence for the multi-view contribution.
  4. [§7.2 and §9] The experimental section does not report training steps, batch size, learning rate, or other training hyperparameters, and Section 9 acknowledges that 'our current model has not finished training' and that hyperparameters 'need to be further fine-tuned.' As a result, the preliminary results cannot be interpreted as a controlled comparison with Zero-1-to-3 or as evidence for any architectural conclusion.
minor comments (6)
  1. [§5.2] The phrase 'making the Q and K projections computationally redundant' is imprecise: the projections are still computed, but their outputs do not affect the attention output when there is a single key.
  2. [Table 2] Table 2 contains only a header row and no generated images or metrics despite the text referring to its contents; either populate the table or remove the reference.
  3. [Throughout] There are numerous typos and formatting issues, including 'tavle' for 'table' in Section 7.3, 'V AEs' for 'VAEs' in Section 1, and 'misinformation and misinformation' in Section 10.
  4. [§2.4 and §3.2] The dataset names 'NeRF4' and 'RealFusion15' are nonstandard and should be defined, and the method is referred to inconsistently as 'Zero-1-to-3++' in Section 2.3 and 'Stable Zero-1-to-3' in Section 3.2.
  5. [§6.3.1 and §7.1.2] The loss function notation 'Concat[{c(xi, Ri, Ti)n i }]' is malformed and should be typeset with proper indices and formatting.
  6. [References] The paper cites Gao et al. (2022) for NeRF in several places but should cite the original formulation of Mildenhall et al. (2021) as well.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity; the attention-collapse claim is a direct softmax identity under the paper's single-token premise, with the main risk being an unreproduced code-inspection premise rather than a circular derivation.

full rationale

The paper's central diagnostic is not a circular prediction. Section 5.2 starts from the standard softmax definition and the premise that Zero-1-to-3 supplies a single 1x768 context vector, from which softmax(QK^T/sqrt(d_k)) over one key is identically 1 and Attention(Q,K,V)=V. That is a direct mathematical consequence of the premise, not a fitted quantity used to predict itself. No parameter is fit to a target and then reported as a prediction; the proposed fixes in Sections 6 and 7 are untested (Section 6.5 postpones implementation; Section 7.3 states 'underwhelming generation quality' and skips metrics), so any claimed improvement is unsupported, but that is a validation gap rather than circular reasoning. The paper does invoke code inspection (Section 5.3, 'line 189') as the bridge from the premise to Zero-1-to-3's actual implementation, and that evidence is not reproduced in the preprint. This makes the load-bearing assumption vulnerable to falsification if the released code passes more than one key, but an unverified empirical premise is not the same as a derivation that reduces to its own inputs. There are no self-citations used as load-bearing support and no imported uniqueness theorem. Accordingly, no specific circular step can be exhibited, and the circularity score is 0.

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

The paper introduces no new physical or mathematical entities. Its central derivation uses a standard property of softmax and an empirical claim about the Zero-1-to-3 codebase that is not shown. The proposed architectural fixes rest on an unproven assumption about the sufficiency of vertical concatenation.

assumptions (3)
  • standard math Softmax over a single element returns 1.
    Used in Section 5.2 to assert that cross-attention output equals V.
  • domain assumption Zero-1-to-3 passes the image and camera condition as a single 1x768 vector to cross-attention.
    Stated in Sections 5.1 and 5.3 based on code inspection; not demonstrated in this preprint.
  • ad hoc to paper The proposed vertical concatenation of image and angle embeddings is sufficient to make cross-attention effective.
    Assumed in Section 7.1 without supporting experiments; the paper reports poor quality.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Fixing the Perspective: A Critical Examination of Zero-1-to-3." pith.science (2026). https://pith.science/paper/WOARI6UM

@misc{pith2026241115706,
  author       = {Pith},
  title        = {Pith review of: Fixing the Perspective: A Critical Examination of Zero-1-to-3},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WOARI6UM}},
  note         = {Machine review of arXiv:2411.15706}
}
read the original abstract

Novel view synthesis is a fundamental challenge in image-to-3D generation, requiring the generation of target view images from a set of conditioning images and their relative poses. While recent approaches like Zero-1-to-3 have demonstrated promising results using conditional latent diffusion models, they face significant challenges in generating consistent and accurate novel views, particularly when handling multiple conditioning images. In this work, we conduct a thorough investigation of Zero-1-to-3's cross-attention mechanism within the Spatial Transformer of the diffusion 2D-conditional UNet. Our analysis reveals a critical discrepancy between Zero-1-to-3's theoretical framework and its implementation, specifically in the processing of image-conditional context. We propose two significant improvements: (1) a corrected implementation that enables effective utilization of the cross-attention mechanism, and (2) an enhanced architecture that can leverage multiple conditional views simultaneously. Our theoretical analysis and preliminary results suggest potential improvements in novel view synthesis consistency and accuracy.

Figures

Figures reproduced from arXiv: 2411.15706 by the authors.

Figure 1
Figure 1. Architectural overview of multi-view condi [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Cross-attention mechanism within a single [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Example of back-view generation artifacts [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Multi-view architecture overview. Multiple in [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Architecture for our proposed fix to Zero-1- [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: The embedding difference between Zero-1-to [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 13 canonical work pages

  1. [1]

    URL: " 'urlintro :=

    ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRINGS urlintro eprinturl eprintpr...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    James F Blinn. 2005. What is a pixel? IEEE computer graphics and applications, 25(5):82--87

    work page 2005

  4. [4]

    Kevin Chen, Christopher B Choy, Manolis Savva, Angel X Chang, Thomas Funkhouser, and Silvio Savarese. 2019. Text2shape: Generating shapes from natural language by learning joint embeddings. In Computer Vision--ACCV 2018: 14th Asian Conference on Computer Vision, Perth, Australia, December 2--6, 2018, Revised Selected Papers, Part III 14, pages 100--116. Springer

    work page 2019

  5. [5]

    Matt Deitke, Dustin Schwenk, Jordi Salvador, Luca Weihs, Oscar Michel, Eli VanderBilt, Ludwig Schmidt, Kiana Ehsani, Aniruddha Kembhavi, and Ali Farhadi. 2022. http://arxiv.org/abs/2212.08051 Objaverse: A universe of annotated 3d objects

    arXiv 2022

  6. [6]

    Kyle Gao, Yina Gao, Hongjie He, Dening Lu, Linlin Xu, and Jonathan Li. 2022. Nerf: Neural radiance field in 3d vision, a comprehensive review. arXiv preprint arXiv:2210.00379

    arXiv 2022

  7. [7]

    Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. 2014. Generative adversarial nets. Advances in neural information processing systems, 27

    2014

  8. [8]

    Yulan Guo, Hanyun Wang, Qingyong Hu, Hao Liu, Li Liu, and Mohammed Bennamoun. 2020. Deep learning for 3d point clouds: A survey. IEEE transactions on pattern analysis and machine intelligence, 43(12):4338--4364

    work page 2020

Show all 35 references
  1. [9]

    Jack Hessel, Ari Holtzman, Maxwell Forbes, Ronan Le Bras, and Yejin Choi. 2021. Clipscore: A reference-free evaluation metric for image captioning. arXiv preprint arXiv:2104.08718

    2021 arXiv

  2. [10]

    Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter. 2018. http://arxiv.org/abs/1706.08500 Gans trained by a two time-scale update rule converge to a local nash equilibrium

    2018 arXiv

  3. [11]

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. 2020. Denoising diffusion probabilistic models. Advances in neural information processing systems, 33:6840--6851

    2020

  4. [12]

    Jonathan Ho and Tim Salimans. 2022. Classifier-free diffusion guidance. arXiv preprint arXiv:2207.12598

    2022 arXiv

  5. [13]

    Ajay Jain, Ben Mildenhall, Jonathan T Barron, Pieter Abbeel, and Ben Poole. 2022. Zero-shot text-guided object generation with dream fields. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 867--876

    2022

  6. [14]

    Diederik P Kingma and Max Welling. 2013. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114

    2013 arXiv

  7. [15]

    Chen-Hsuan Lin, Jun Gao, Luming Tang, Towaki Takikawa, Xiaohui Zeng, Xun Huang, Karsten Kreis, Sanja Fidler, Ming-Yu Liu, and Tsung-Yi Lin. 2023. Magic3d: High-resolution text-to-3d content creation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Reco...

    2023

  8. [16]

    Ruoshi Liu, Rundi Wu, Basile Van Hoorick, Pavel Tokmakov, Sergey Zakharov, and Carl Vondrick. 2023. http://arxiv.org/abs/2303.11328 Zero-1-to-3: Zero-shot one image to 3d object

    2023 arXiv

  9. [17]

    Luke Melas-Kyriazi, Christian Rupprecht, Iro Laina, and Andrea Vedaldi. 2023. https://arxiv.org/abs/2302.10663 Realfusion: 360 reconstruction of any object from a single image . In CVPR

    2023 arXiv

  10. [18]

    Ben Mildenhall, Pratul P Srinivasan, Matthew Tancik, Jonathan T Barron, Ravi Ramamoorthi, and Ren Ng. 2021. Nerf: Representing scenes as neural radiance fields for view synthesis. Communications of the ACM, 65(1):99--106

    2021

  11. [19]

    Shentong Mo, Enze Xie, Ruihang Chu, Lewei Yao, Lanqing Hong, Matthias Nießner, and Zhenguo Li. 2023. Dit-3d: Exploring plain diffusion transformers for 3d shape generation. arXiv preprint arXiv: 2307.01831

    2023 arXiv

  12. [20]

    Nasir Mohammad Khalid, Tianhao Xie, Eugene Belilovsky, and Tiberiu Popa. 2022. Clip-mesh: Generating textured meshes from text using pretrained image-text models. In SIGGRAPH Asia 2022 conference papers, pages 1--8

    2022

  13. [21]

    Alex Nichol, Prafulla Dhariwal, Aditya Ramesh, Pranav Shyam, Pamela Mishkin, Bob McGrew, Ilya Sutskever, and Mark Chen. 2021. Glide: Towards photorealistic image generation and editing with text-guided diffusion models. arXiv preprint arXiv:2112.10741

    2021 arXiv

  14. [22]

    Barron, and Ben Mildenhall

    Ben Poole, Ajay Jain, Jonathan T. Barron, and Ben Mildenhall. 2022. http://arxiv.org/abs/2209.14988 Dreamfusion: Text-to-3d using 2d diffusion

    2022 arXiv

  15. [23]

    Guocheng Qian, Jinjie Mai, Abdullah Hamdi, Jian Ren, Aliaksandr Siarohin, Bing Li, Hsin-Ying Lee, Ivan Skorokhodov, Peter Wonka, Sergey Tulyakov, and Bernard Ghanem. 2023. http://arxiv.org/abs/2306.17843 Magic123: One image to high-quality 3d object generation using both 2d an...

    2023 arXiv

  16. [24]

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, and Ilya Sutskever. 2021. http://arxiv.org/abs/2103.00020 Learning transferable visual models from natural lang...

    2021 arXiv

  17. [25]

    Aditya Ramesh, Prafulla Dhariwal, Alex Nichol, Casey Chu, and Mark Chen. 2022. Hierarchical text-conditional image generation with clip latents. arXiv preprint arXiv:2204.06125, 1(2):3

    2022 arXiv

  18. [26]

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj \"o rn Ommer. 2022 a . High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10684--10695

    2022

  19. [27]

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj\"orn Ommer. 2022 b . High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 10684--10695

    2022

  20. [28]

    Aditya Sanghi, Hang Chu, Joseph G Lambourne, Ye Wang, Chin-Yi Cheng, Marco Fumero, and Kamal Rahimi Malekshan. 2022. Clip-forge: Towards zero-shot text-to-shape generation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 18603--18613

    2022

  21. [29]

    Hoigi Seo, Hayeon Kim, Gwanghyun Kim, and Se Young Chun. 2023 a . Ditto-nerf: Diffusion-based iterative text to omni-directional 3d model. arXiv preprint arXiv:2304.02827

    2023 arXiv

  22. [30]

    Junyoung Seo, Wooseok Jang, Min-Seop Kwak, Hyeonsu Kim, Jaehoon Ko, Junho Kim, Jin-Hwa Kim, Jiyoung Lee, and Seungryong Kim. 2023 b . Let 2d diffusion model know 3d-consistency for robust text-to-3d generation. arXiv preprint arXiv:2303.07937

    2023 arXiv

  23. [31]

    Ruoxi Shi, Hansheng Chen, Zhuoyang Zhang, Minghua Liu, Chao Xu, Xinyue Wei, Linghao Chen, Chong Zeng, and Hao Su. 2023. Zero123++: a single image to consistent multi-view diffusion base model. arXiv preprint arXiv:2310.15110

    2023 arXiv

  24. [32]

    Yang Song, Jascha Sohl-Dickstein, Diederik P Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole. 2020. Score-based generative modeling through stochastic differential equations. arXiv preprint arXiv:2011.13456

    2020 arXiv

  25. [33]

    Can Wang, Menglei Chai, Mingming He, Dongdong Chen, and Jing Liao. 2022. Clip-nerf: Text-and-image driven manipulation of neural radiance fields. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 3835--3844

    2022

  26. [34]

    He Wang and Juyong Zhang. 2022. A survey of deep learning-based mesh processing. Communications in Mathematics and Statistics, 10(1):163--194

    2022

  27. [35]

    Efros, Eli Shechtman, and Oliver Wang

    Richard Zhang, Phillip Isola, Alexei A. Efros, Eli Shechtman, and Oliver Wang. 2018. http://arxiv.org/abs/1801.03924 The unreasonable effectiveness of deep features as a perceptual metric

    2018 arXiv

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.