Pith. sign in

REVIEW 5 major objections 6 minor 49 references

Text-guided Synthetic Geometric Augmentation for Zero-shot 3D Understanding

T0 review · 5 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper claims that synthetic 3D point clouds generated by Point-E and filtered by text-image consistency can be added to ShapeNet to improve zero-shot 3D classification, with gains of 3.0% on Objaverse-LVIS, 4.6% on ScanObjectNN, and…

desk verdict A sensible idea with an internally inconsistent main result; the headline claim that synthetic expansion helps 3D zero-shot classification needs more careful experiments before it can be trusted. read the letter →

arxiv 2501.09278 v2 pith:FO7ZZEXZ submitted 2025-01-16 cs.CV

Kohei Torimi , Ryosuke Yamada , Daichi Otsuka , Kensho Hara , Yuki M. Asano , Hirokatsu Kataoka , Yoshimitsu Aoki This is my paper · ORCID
classification cs.CV
keywords syntheticdatazero-shot3Dclassificationtext-to-3DpointcloudconsistencyfilteringcontrastivelearningTeGAdatasetexpansion
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 show that synthetic 3D data generated by a text-to-3D model can serve as effective training data for zero-shot 3D classification, alleviating the high cost of collecting real 3D data. It proposes TeGA, which generates point clouds from class names using Point-E, renders them into images, and filters out samples whose images do not match the source text. When these filtered synthetic samples are added to ShapeNet, the MixCon3D model improves by 3.0% on Objaverse-LVIS, 4.6% on ScanObjectNN, and 8.7% on ModelNet40 compared to training on ShapeNet alone. This matters because it suggests a practical route to scaling 3D training data without manual annotation, addressing a core bottleneck in zero-shot 3D vision.

What carries the argument

The key machinery is TeGA's consistency filtering pipeline. For each generated point cloud, the method renders two images (front and back), generates captions with BLIP, merges them via GPT-4 into one unified caption, and then scores the match between this caption and the original text prompt using two metrics: word-level containment (score 5 or 1) and a GPT-4 semantic similarity score (1 to 5). The scores are summed and compared to a threshold (δ = 3.5); samples below the threshold are discarded. This filter is the component that keeps synthetic data aligned across text, image, and point cloud modalities during contrastive pretraining.

What would settle it

A direct test would be to measure the alignment between the generated point clouds and the input text using a point-cloud-caption retrieval model, and compare alignment scores for samples that pass TeGA's image-text filter versus those that fail. If the point-cloud-text alignment is not higher for filtered samples, the filter's core assumption fails. A second check would be to train MixCon3D using unfiltered synthetic data on ModelNet40 alone and see whether accuracy remains above the filtered result, which would indicate that filtering removes useful geometric diversity in that setting.

Watch

Extended reading notes

Core claim

The central claim is that a consistency-filtered synthetic dataset, constructed by feeding ShapeNet category names into Point-E and rendering the resulting point clouds from multiple viewpoints, can be combined with a real 3D dataset to improve zero-shot 3D classification performance. The authors show that doubling ShapeNet with TeGA-filtered synthetic data improves MixCon3D's Top-1 accuracy on three benchmarks: from 9.4% to 12.4% on Objaverse-LVIS, from 46.5% to 51.1% on ScanObjectNN, and from 64.6% to 73.3% on ModelNet40. They further show that the consistency filter, which removes roughly half of the generated samples, is beneficial on Objaverse-LVIS and ScanObjectNN but slightly harmful on ModelNet40, and that replacing 25% of real ShapeNet data with synthetic data beats ShapeNet alone while 100% synthetic data collapses accuracy.

Load-bearing premise

The consistency filter assumes that if the input text matches the final rendered image, the intermediate point cloud also matches the text, even though the point cloud itself is never directly checked.

Editorial extensions

If this is right

  • Doubling ShapeNet with TeGA-filtered Point-E data improves zero-shot classification accuracy on Objaverse-LVIS, ScanObjectNN, and ModelNet40.
  • Consistency filtering removes noisy samples and improves accuracy on Objaverse-LVIS and ScanObjectNN, though it slightly reduces ModelNet40 accuracy.
  • Replacing 25% of real ShapeNet data with synthetic data outperforms ShapeNet alone, while training on 100% synthetic data leads to dramatic performance collapse.
  • Scaling synthetic data to twice the ShapeNet size improves Objaverse-LVIS and ModelNet40 but degrades ScanObjectNN, which is more sensitive to noise.
  • The text-image alignment filter is sufficient to support modality alignment in language-image-3D contrastive learning, even without direct point-cloud verification.

Reading between the lines

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

  • The observed gains may stem more from increased data volume and stronger reliance on text-image alignment than from geometric fidelity, since the paper's own feature visualization shows that real and synthetic features of the same class remain separated.
  • The filter's assumption that text-image alignment implies point-cloud alignment is directly testable; if it fails, a point-cloud-aware alignment check could recover the ModelNet40 accuracy lost by filtering.
  • Applying TeGA to a larger and more diverse real dataset (e.g., the combined ShapeNet, ABO, 3D-FUTURE, and Objaverse setup used by state-of-the-art methods) could reveal whether synthetic expansion remains beneficial at higher data volumes or whether the domain gap between Point-E and real shapes becomes a limiting factor.
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

5 major / 6 minor

Summary. The paper proposes TeGA, a method for expanding a 3D training dataset by generating synthetic point clouds with the text-to-3D model Point-E, rendering multi-view images from the generated point clouds, and applying a consistency filter that uses BLIP captions and GPT-4 scoring to remove samples whose text and rendered image appear misaligned. The augmented dataset is then used to train MixCon3D, a language-image-3D contrastive model, and the central claim is that training on ShapeNet plus the TeGA-filtered synthetic data improves zero-shot 3D classification on Objaverse-LVIS, ScanObjectNN, and ModelNet40 relative to training on ShapeNet alone. The main evidence is reported in Table 3, with supporting ablations on filtering, guidance scale, mixing ratio, and scaling in Tables 1, 2, 4, and 5.

Significance. If the reported improvements are robust, the paper would provide a practical and timely demonstration that off-the-shelf text-to-3D generators can mitigate the data-scarcity bottleneck in zero-shot 3D classification, and the proposed consistency filtering is a sensible safeguard against misaligned synthetic samples. The paper is also useful in that it reports ablations of several important design choices. However, the central quantitative claim is not reproducible from the experiments as reported: the scaling ablation in Table 5 contradicts the headline results in Table 3, the filtering ablation in Table 1 shows a large negative effect on ModelNet40, and no error bars, seeds, or code are provided. These issues prevent the reader from assessing whether the claimed gains are reliable or are artifacts of a particular run or an unreported configuration.

major comments (5)
  1. [§5.3 vs. §5.4, Tables 3 and 5] The headline result in Table 3 (ShapeNet + TeGA at 12.4/51.1/73.3 on Objaverse-LVIS, ScanObjectNN, ModelNet40) is not consistent with the scaling ablation in Table 5. The ×1 row of Table 5 is the same nominal configuration (ShapeNet plus Point-E data equal to ShapeNet's size), yet it reports 11.2/47.1/73.4. The gap is 1.2 points on Objaverse-LVIS (40% of the claimed 3.0-point gain) and 4.0 points on ScanObjectNN (87% of the claimed 4.6-point gain). Section 5.4 does not state whether the Table 5 runs use the consistency filter; if they do, the two tables are irreconcilable, and if they do not, the paper fails to identify which component produced the Table 3 results. The authors need to clarify the exact configuration and report multiple seeds with standard deviations.
  2. [§5.2, Table 1] The consistency filtering ablation contradicts the paper's claim that filtering is uniformly beneficial. Using 10,000 synthetic samples, filtering improves Objaverse-LVIS and ScanObjectNN but decreases ModelNet40 by 2.1 points (73.4 without filtering vs. 71.3 with filtering). Section 5.2 states that the results 'indicate that the proposed consistency filtering of TeGA provides effective synthetic training,' which is only partially supported. Since the main result in Table 3 uses the filter at a much larger scale, the paper should explain why filtering helps at the scale used in Table 3, or qualify the claim if the effect is benchmark-dependent.
  3. [§5.4, Table 5] Table 5 shows that ScanObjectNN accuracy declines monotonically with the amount of synthetic data (48.2 at ×0.1, 47.1 at ×1, 45.1 at ×2), which is in direct tension with Table 3, where the ×1 setting yields 51.1. The paper acknowledges this decline in the text, but does not reconcile it with the large ScanObjectNN gain claimed in the main experiment. The authors need to state whether the Table 5 runs used the same filtered dataset as Table 3, and if so, explain why the identical configuration yields such different numbers.
  4. [§4, Consistency Filtering; Appendix Figure C] The consistency filter assumes that alignment between the input text and the final rendered image implies alignment of the intermediate point cloud. This is load-bearing because the point cloud is never checked directly. The paper's own feature visualization in Appendix Figure C shows that real and synthetic features of the same class are well separated, which indicates a substantial domain gap between ShapeNet and Point-E data. The authors should discuss whether the text-image-based filter is sufficient to ensure point-cloud quality, or whether the domain gap explains why synthetic data sometimes hurts performance (e.g., ModelNet40 in Table 1 and ScanObjectNN in Table 5).
  5. [All experiments] No error bars, number of seeds, or code are reported for any of the tables. Given the contradictions between Tables 3 and 5 and the negative scaling trend in Table 5, single-run numbers are insufficient to support the paper's central claim. Reporting mean and standard deviation over at least three seeds, or releasing the code and generated dataset, is necessary to establish that the gains are not due to run-to-run variance.
minor comments (6)
  1. [Abstract and §5.3] The abstract reports gains of 4.6% on ScanObjectNN and 8.7% on ModelNet40, while Section 5.3 reports 4.4% and 8.8% for the same comparison; please harmonize these numbers.
  2. [§1 and §5.4] The size of ShapeNet is given as 52,470 in the Introduction and 53,470 in Section 5.4; please correct the inconsistency.
  3. [§4, Synthetic Dataset Construction] The notation in the definition of the synthetic dataset is confusing: the text refers to Dt and Ds in ways that appear swapped, and the tuple contains a typo 'x0T′ i'. Please clarify which dataset is the source and which is the generated one.
  4. [§5.2, Effect of Guidance Scale] The text says 'In Table 1, we compare the Top1 accuracy with each guidance scale' but the results are in Table 2; please fix the cross-reference.
  5. [§5.4, Figure 5 caption] The caption of Figure 5 refers to 'PE/SP' but the table and text use 'PE/SN'; please use a consistent abbreviation.
  6. [Abstract] The phrase 'which achieves SoTA in zero-shot 3D classification' is unclear because it could refer to TeGA itself or to the base MixCon3D model; please rephrase to avoid ambiguity.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: TeGA's claimed gains are an empirical comparison on held-out zero-shot benchmarks, and no fitted parameter or self-citation is load-bearing.

full rationale

The paper's derivation chain is empirical rather than definitional. Synthetic point clouds are produced by Point-E (Eq. 3), rendered (Eq. 4), filtered by text-image consistency (Eqs. 5-7), and used to train MixCon3D with the contrastive objective (Eq. 2). None of these equations consumes Objaverse-LVIS, ScanObjectNN, or ModelNet40 labels; the consistency threshold is set on ShapeNet ('We experimentally validated this threshold δ during data generation with ShapeNet and finally set it to 3.5'), so the headline numbers are not fitted to the target test sets. The inference that text-image alignment implies point-cloud alignment ('If alignment is maintained between the input text xT_i and the final output image xI'_i, it can be inferred that the intermediate output, the point cloud xP'_i, is also aligned') is a heuristic assumption, and the paper's own Figure C shows real and synthetic features of the same class are separated, but this is a limitation of the filter, not a circular reduction of the benchmark result to its inputs. References to Point-E, MixCon3D, BLIP, GPT-4, and OpenCLIP are external prior work, not self-citations that carry the argument. The internal inconsistency between Table 3 and the ×1 row of Table 5 is a reproducibility and correctness concern, but it does not make the derivation circular.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The method relies on a handful of domain assumptions about generative model fidelity and filtering reliability, plus two hand-tuned parameters (delta and guidance scale). No new physical entities are introduced.

free parameters (3)
  • consistency filtering threshold delta = 3.5
    Set after 'experimentally validated during data generation with ShapeNet' (Sec 4). Determines how much synthetic data is kept; sensitivity not reported.
  • Point-E guidance scale omega = [3.0, 0.0]
    Chosen from Table 2 experiments over {0.3, 3.0, 30}; 3.0 yields best downstream performance. A tuning choice, not a target fit.
  • Number of rendering viewpoints = 20
    Fixed design choice (18-degree rotations) for generating multi-view images; no ablation reported.
assumptions (5)
  • domain assumption If alignment holds between input text and rendered image, the intermediate point cloud is also aligned.
    States in Sec 4, Consistency Filtering: 'If alignment is maintained between the input text xT_i and the final output image xI'_i, it can be inferred that the intermediate output, the point cloud xP'_i, is also aligned.' Figure C shows real and synthetic features for the same class are separated, weakening this assumption.
  • domain assumption Point-E generated point clouds are semantically faithful to ShapeNet class-name prompts.
    The whole method depends on Point-E producing shapes that match the category text. The paper acknowledges failures and uses filtering, but the surviving data still has a domain gap (Figure C).
  • domain assumption Ball Pivoting Algorithm meshing of Point-E point clouds produces renderable surfaces.
    Sec 4: Point-E point clouds are 'not designed for meshing', and the authors adopt BPA because state-of-the-art meshing collapses. No quantitative evaluation of mesh quality is provided.
  • domain assumption BLIP and GPT-4 provide reliable semantic alignment scores for filtering.
    The consistency filter relies on BLIP captions and a GPT-4 five-point score; the GPT-4 prompt is redesigned from T3Bench and word matching is added to compensate for GPT-4 failures (Appendix A.2). The reliability of these scores is not validated.
  • standard math The contrastive loss L_All (Eq. 2) is a valid objective for aligning image, text, and point cloud embeddings.
    Standard contrastive learning formulation from MixCon3D; not novel.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Text-guided Synthetic Geometric Augmentation for Zero-shot 3D Understanding." pith.science (2026). https://pith.science/paper/FO7ZZEXZ

@misc{pith2026250109278,
  author       = {Pith},
  title        = {Pith review of: Text-guided Synthetic Geometric Augmentation for Zero-shot 3D Understanding},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FO7ZZEXZ}},
  note         = {Machine review of arXiv:2501.09278}
}
read the original abstract

Zero-shot recognition models require extensive training data for generalization. However, in zero-shot 3D classification, collecting 3D data and captions is costly and laborintensive, posing a significant barrier compared to 2D vision. Recent advances in generative models have achieved unprecedented realism in synthetic data production, and recent research shows the potential for using generated data as training data. Here, naturally raising the question: Can synthetic 3D data generated by generative models be used as expanding limited 3D datasets? In response, we present a synthetic 3D dataset expansion method, Textguided Geometric Augmentation (TeGA). TeGA is tailored for language-image-3D pretraining, which achieves SoTA in zero-shot 3D classification, and uses a generative textto-3D model to enhance and extend limited 3D datasets. Specifically, we automatically generate text-guided synthetic 3D data and introduce a consistency filtering strategy to discard noisy samples where semantics and geometric shapes do not match with text. In the experiment to double the original dataset size using TeGA, our approach demonstrates improvements over the baselines, achieving zeroshot performance gains of 3.0% on Objaverse-LVIS, 4.6% on ScanObjectNN, and 8.7% on ModelNet40. These results demonstrate that TeGA effectively bridges the 3D data gap, enabling robust zero-shot 3D classification even with limited real training data and paving the way for zero-shot 3D vision application.

Figures

Figures reproduced from arXiv: 2501.09278 by the authors.

Figure 1
Figure 1. Our proposed TeGA (Text-guided Geometric Augmen [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. A visualization of synthetic 3D data generated from real 3D data and Point-E. Synthetic 3D data shows that it is more difficult to generate detailed geometrical detail compared to real data. embeddings with the rich, pre-trained feature spaces of im￾ages and text to facilitate the classification of 3D data in￾cluding unseen classes. This facilitation enables language￾image-3D contrastive learning to mitigate the imp… view at source ↗
Figure 4
Figure 4. The overview of consistency filtering process. The purpose of this process is to remove misaligned data that may introduce [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: Confusion matrics when varying PE/SP. When the proportion of real data decreases, the model’s predictions increasingly skew [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 29 canonical work pages

  1. [1]

    Visual objects in context

    Moshe Bar. Visual objects in context. Nature Reviews Neu- roscience, 5(8):617–629, 2004. 3

    work page 2004

  2. [2]

    The ball-pivoting algo- rithm for surface reconstruction

    Fausto Bernardini, Joshua Mittleman, Holly Rushmeier, Claudio Silva, and Gabriel Taubin. The ball-pivoting algo- rithm for surface reconstruction. IEEE Transactions on Vi- sualization and Computer Graphics, 5(4):349–359, 1999. 5

    work page 1999

  3. [3]

    Shapenet: An information-rich 3d model repository

    Angel X Chang, Thomas Funkhouser, Leonidas Guibas, Pat Hanrahan, Qixing Huang, Zimo Li, Silvio Savarese, Manolis Savva, Shuran Song, Hao Su, et al. Shapenet: An information-rich 3d model repository. arXiv preprint arXiv:1512.03012, 2015. 6

    arXiv 2015

  4. [4]

    Fan- tasia3d: Disentangling geometry and appearance for high- quality text-to-3d content creation

    Rui Chen, Yongwei Chen, Ningxin Jiao, and Kui Jia. Fan- tasia3d: Disentangling geometry and appearance for high- quality text-to-3d content creation. In Proceedings of the IEEE/CVF international conference on computer vision , pages 22246–22256, 2023. 2, 3

    work page 2023

  5. [5]

    Reproducible scal- ing laws for contrastive language-image learning

    Mehdi Cherti, Romain Beaumont, Ross Wightman, Mitchell Wortsman, Gabriel Ilharco, Cade Gordon, Christoph Schuh- mann, Ludwig Schmidt, and Jenia Jitsev. Reproducible scal- ing laws for contrastive language-image learning. In Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 2818–2829, 2023. 3, 6, 11

    work page 2023

  6. [6]

    Abo: Dataset and benchmarks for real-world 3d object un- derstanding

    Jasmine Collins, Shubham Goel, Kenan Deng, Achlesh- war Luthra, Leon Xu, Erhan Gundogdu, Xi Zhang, Tomas F Yago Vicente, Thomas Dideriksen, Himanshu Arora, et al. Abo: Dataset and benchmarks for real-world 3d object un- derstanding. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 21126– 21136, 2022. 6

    work page 2022

  7. [7]

    Objaverse: A universe of annotated 3d objects

    Matt Deitke, Dustin Schwenk, Jordi Salvador, Luca Weihs, Oscar Michel, Eli VanderBilt, Ludwig Schmidt, Kiana Ehsani, Aniruddha Kembhavi, and Ali Farhadi. Objaverse: A universe of annotated 3d objects. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR), pages 13142–13153, 2023. 2, 6

    work page 2023

  8. [8]

    Objaverse-xl: A universe of 10m+ 3d objects

    Matt Deitke, Ruoshi Liu, Matthew Wallingford, Huong Ngo, Oscar Michel, Aditya Kusupati, Alan Fan, Christian Laforte, Vikram V oleti, Samir Yitzhak Gadre, et al. Objaverse-xl: A universe of 10m+ 3d objects. Advances in Neural Informa- tion Processing Systems, 36, 2024. 2

    2024

Show all 49 references
  1. [9]

    3d-future: 3d fur- niture shape with texture

    Huan Fu, Rongfei Jia, Lin Gao, Mingming Gong, Binqiang Zhao, Steve Maybank, and Dacheng Tao. 3d-future: 3d fur- niture shape with texture. International Journal of Computer Vision, 129:3313–3337, 2021. 6

    2021

  2. [10]

    Sculpting holistic 3d representation in con- trastive language-image-3d pre-training

    Yipeng Gao, Zeyu Wang, Wei-Shi Zheng, Cihang Xie, and Yuyin Zhou. Sculpting holistic 3d representation in con- trastive language-image-3d pre-training. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR), pages 22998–23008, 2024. 2, 3

    2024

  3. [11]

    Generative adversarial networks

    Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial networks. Commu- nications of the ACM, 63(11):139–144, 2020. 3

    2020

  4. [12]

    T3 bench: Benchmarking current progress in text-to-3d gener- ation

    Yuze He, Yushi Bai, Matthieu Lin, Wang Zhao, Yubin Hu, Jenny Sheng, Ran Yi, Juanzi Li, and Yong-Jin Liu. T3 bench: Benchmarking current progress in text-to-3d gener- ation. arXiv preprint arXiv:2310.02977, 2023. 5, 6, 11

    2023 arXiv

  5. [13]

    Clip goes 3d: Leveraging prompt tuning for language grounded 3d recognition

    Deepti Hegde, Jeya Maria Jose Valanarasu, and Vishal Pa- tel. Clip goes 3d: Leveraging prompt tuning for language grounded 3d recognition. In Proceedings of the IEEE/CVF International Conference on Computer Vision(CVPR), pages 2028–2038, 2023. 2, 3

    2023

  6. [14]

    Clip2point: Transfer clip to point cloud classifi- cation with image-depth pre-training

    Tianyu Huang, Bowen Dong, Yunhan Yang, Xiaoshui Huang, Rynson WH Lau, Wanli Ouyang, and Wangmeng Zuo. Clip2point: Transfer clip to point cloud classifi- cation with image-depth pre-training. In Proceedings of the IEEE/CVF International Conference on Computer Vi- sion(CVPR), pa...

    2023

  7. [15]

    When faking your data actually helps – learn- ing vision from GANs, NeRFs, and noise, 2022

    Phillip Isola. When faking your data actually helps – learn- ing vision from GANs, NeRFs, and noise, 2022. BMVC Keynote talk. 3

    2022

  8. [16]

    Shap-e: Generat- ing conditional 3d implicit functions

    Heewoo Jun and Alex Nichol. Shap-e: Generat- ing conditional 3d implicit functions. arXiv preprint arXiv:2305.02463, 2023. 2

    2023 arXiv

  9. [17]

    Object category structure in response patterns of neuronal population in monkey inferior temporal cortex

    Roozbeh Kiani, Hossein Esteky, Koorosh Mirpour, and Keiji Tanaka. Object category structure in response patterns of neuronal population in monkey inferior temporal cortex. Journal of neurophysiology, 97(6):4296–4309, 2007. 3

    2007

  10. [18]

    Auto-encoding variational bayes

    Diederik P Kingma. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114, 2013. 3

    2013 arXiv

  11. [19]

    Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models

    Junnan Li, Dongxu Li, Silvio Savarese, and Steven Hoi. Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models. In In- ternational conference on machine learning , pages 19730– 19742. PMLR, 2023. 5

    2023

  12. [20]

    Magic3d: High-resolution text-to-3d content creation

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

    2023

  13. [21]

    Openshape: Scaling up 3d shape representation towards open-world understanding

    Minghua Liu, Ruoxi Shi, Kaiming Kuang, Yinhao Zhu, Xu- anlin Li, Shizhong Han, Hong Cai, Fatih Porikli, and Hao Su. Openshape: Scaling up 3d shape representation towards open-world understanding. Advances in Neural Information Processing Systems, 36, 2024. 2, 3, 6

    2024

  14. [22]

    Zero-1-to- 3: Zero-shot one image to 3d object

    Ruoshi Liu, Rundi Wu, Basile Van Hoorick, Pavel Tok- makov, Sergey Zakharov, and Carl V ondrick. Zero-1-to- 3: Zero-shot one image to 3d object. In Proceedings of the IEEE/CVF international conference on computer vision, pages 9298–9309, 2023. 2

    2023

  15. [23]

    Decoupled weight decay regularization

    I Loshchilov. Decoupled weight decay regularization. arXiv preprint arXiv:1711.05101, 2017. 6

    2017 arXiv

  16. [24]

    Sgdr: Stochas- tic gradient descent with warm restarts

    Ilya Loshchilov and Frank Hutter. Sgdr: Stochas- tic gradient descent with warm restarts. arXiv preprint arXiv:1608.03983, 2016. 6

    2016 arXiv

  17. [25]

    Glide: Towards photorealistic image generation and editing with text-guided diffusion models.arXiv preprint arXiv:2112.10741, 2021

    Alex Nichol, Prafulla Dhariwal, Aditya Ramesh, Pranav Shyam, Pamela Mishkin, Bob McGrew, Ilya Sutskever, and Mark Chen. Glide: Towards photorealistic image generation and editing with text-guided diffusion models.arXiv preprint arXiv:2112.10741, 2021. 3

    2021 arXiv

  18. [26]

    Point-e: A system for generat- ing 3d point clouds from complex prompts

    Alex Nichol, Heewoo Jun, Prafulla Dhariwal, Pamela Mishkin, and Mark Chen. Point-e: A system for generat- ing 3d point clouds from complex prompts. arXiv preprint arXiv:2212.08751, 2022. 2, 3, 4

    2022 arXiv

  19. [27]

    Dinov2: Learning robust visual features without supervision

    Maxime Oquab, Timothee Darcet, Theo Moutakanni, Huy V o, Marc Szafraniec, Vasil Khalidov, Pierre Fernandez, Daniel Haziza, Francisco Massa, Alaaeldin El-Nouby, et al. Dinov2: Learning robust visual features without supervision. arXiv preprint arXiv:2304.07193, 2023. 3

    2023 arXiv

  20. [28]

    Dreamfusion: Text-to-3d using 2d diffusion

    Ben Poole, Ajay Jain, Jonathan T Barron, and Ben Milden- hall. Dreamfusion: Text-to-3d using 2d diffusion. arXiv preprint arXiv:2209.14988, 2022. 2, 3

    2022 arXiv

  21. [29]

    Contrast with reconstruct: Contrastive 3d representation learning guided by generative pretraining

    Zekun Qi, Runpei Dong, Guofan Fan, Zheng Ge, Xiangyu Zhang, Kaisheng Ma, and Li Yi. Contrast with reconstruct: Contrastive 3d representation learning guided by generative pretraining. In International Conference on Machine Learn- ing(ICML), pages 28223–28243. PMLR, 2023. 2, 3

    2023

  22. [30]

    Learn- ing transferable visual models from natural language su- pervision

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learn- ing transferable visual models from natural language su- pervision. In International conference on machine learn- i...

    2021

  23. [31]

    High-resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bjorn Ommer. 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. 3

    2022

  24. [32]

    Photorealistic text-to-image diffusion models with deep language understanding

    Chitwan Saharia, William Chan, Saurabh Saxena, Lala Li, Jay Whang, Emily L Denton, Kamyar Ghasemipour, Raphael Gontijo Lopes, Burcu Karagol Ayan, Tim Salimans, et al. Photorealistic text-to-image diffusion models with deep language understanding. Advances in neural information...

    2022

  25. [33]

    Fake it till you make it: Learning trans- ferable representations from synthetic imagenet clones

    Mert Bulent Sariyildiz, Karteek Alahari, Diane Larlus, and Yannis Kalantidis. Fake it till you make it: Learning trans- ferable representations from synthetic imagenet clones. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 8011–80...

    2023

  26. [34]

    Laion-5b: An open large-scale dataset for training next generation image-text models

    Christoph Schuhmann, Romain Beaumont, Richard Vencu, Cade Gordon, Ross Wightman, Mehdi Cherti, Theo Coombes, Aarush Katta, Clayton Mullis, Mitchell Worts- man, et al. Laion-5b: An open large-scale dataset for training next generation image-text models. Advances in Neural In- f...

    2022

  27. [35]

    Learning vision from mod- els rivals learning vision from data

    Yonglong Tian, Lijie Fan, Kaifeng Chen, Dina Katabi, Dilip Krishnan, and Phillip Isola. Learning vision from mod- els rivals learning vision from data. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR), pages 15887–15898, 2024. 3

    2024

  28. [36]

    Stablerep: Synthetic images from text-to- image models make strong visual representation learners

    Yonglong Tian, Lijie Fan, Phillip Isola, Huiwen Chang, and Dilip Krishnan. Stablerep: Synthetic images from text-to- image models make strong visual representation learners. Advances in Neural Information Processing Systems , 36,

  29. [37]

    Textmesh: Gen- eration of realistic 3d meshes from text prompts

    Christina Tsalicoglou, Fabian Manhardt, Alessio Tonioni, Michael Niemeyer, and Federico Tombari. Textmesh: Gen- eration of realistic 3d meshes from text prompts. In 2024 International Conference on 3D Vision (3DV), pages 1554–

    2024

  30. [38]

    Revisiting point cloud classification: A new benchmark dataset and classification model on real-world data

    Mikaela Angelina Uy, Quang-Hieu Pham, Binh-Son Hua, Thanh Nguyen, and Sai-Kit Yeung. Revisiting point cloud classification: A new benchmark dataset and classification model on real-world data. In Proceedings of the IEEE/CVF International Conference on computer vision(CVPR), pa...

    2019

  31. [39]

    Prolificdreamer: High-fidelity and diverse text-to-3d generation with variational score distilla- tion

    Zhengyi Wang, Cheng Lu, Yikai Wang, Fan Bao, Chongxuan Li, Hang Su, and Jun Zhu. Prolificdreamer: High-fidelity and diverse text-to-3d generation with variational score distilla- tion. Advances in Neural Information Processing Systems , 36, 2024. 2

    2024

  32. [40]

    3d shapenets: A deep representation for volumetric shapes

    Zhirong Wu, Shuran Song, Aditya Khosla, Fisher Yu, Lin- guang Zhang, Xiaoou Tang, and Jianxiong Xiao. 3d shapenets: A deep representation for volumetric shapes. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition(CVPR), pages 1912–1920, 2015. 2, 6

    1912

  33. [41]

    Ulip-2: Towards scalable multimodal pre-training for 3d understanding

    Le Xue, Ning Yu, Shu Zhang, Artemis Panagopoulou, Jun- nan Li, Roberto Martin-Martjn, Jiajun Wu, Caiming Xiong, Ran Xu, Juan Carlos Niebles, et al. Ulip-2: Towards scalable multimodal pre-training for 3d understanding. In Proceed- ings of the IEEE/CVF Conference on Computer Vi...

    2024

  34. [42]

    Uni3d: Exploring unified 3d representation at scale

    Junsheng Zhou, Jinsheng Wang, Baorui Ma, Yu-Shen Liu, Tiejun Huang, and Xinlong Wang. Uni3d: Exploring unified 3d representation at scale. arXiv preprint arXiv:2310.06773,

    arXiv

  35. [47]

    prompt: car prediction: A 3D rendering of a car with a pink and white exterior and a pink interior with red streaks Answer: 5 Word matching 5 Total 10 (safe) Table B

    since the prompt is at the word level, it is inevitable that some detailed information is missing, so exclude it from the point deduction. prompt: car prediction: A 3D rendering of a car with a pink and white exterior and a pink interior with red streaks Answer: 5 Word matchin...

  36. [51]

    prompt: sofa prediction: A modern, cream-colored sofa GPT-4 Answer 1 Word matching 5 Total 6 (safe) Table C

    since the prompt is at the word level, it is inevitable that some detailed information is missing, so exclude it from the point deduction. prompt: sofa prediction: A modern, cream-colored sofa GPT-4 Answer 1 Word matching 5 Total 6 (safe) Table C. The process of consistency fi...

  37. [52]

    If the prediction contains information that does not appear in the prompt, it should not be considered as bad

    Evaluate the recall, or how well the prediction covers the information in the prompt. If the prediction contains information that does not appear in the prompt, it should not be considered as bad

  38. [53]

    Do not provide a complete answer; give the score in the format: 3

    Assign a score between 1 and 5, with 5 being the highest. Do not provide a complete answer; give the score in the format: 3

  39. [54]

    similar in appearance)

    add points if the prediction and prompt are conceptually close (e.g. similar in appearance). (e.g., bike and bycicle and table and chair are close)

  40. [55]

    prompt: birdhouse prediction: A black and white artistic object GPT-4 Answer 2 Word matching 1 Total 3 (ng) Table D

    since the prompt is at the word level, it is inevitable that some detailed information is missing, so exclude it from the point deduction. prompt: birdhouse prediction: A black and white artistic object GPT-4 Answer 2 Word matching 1 Total 3 (ng) Table D. The process of consis...

  41. [2023]

    Details of TeGA A.1

    3 A. Details of TeGA A.1. Details of the generated data by TeGA We provide samples of data generated by TeGA (Fig. A and Fig. B). The generated data which pass consistency filter- ing are shown in Fig. A and the generated data which are filtered out are shown in Fig. B. Tab. A...

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.