pith. sign in

arxiv: 2605.19876 · v1 · pith:OYP74K7Qnew · submitted 2026-05-19 · 💻 cs.CV

Structural Energy Guidance for View-Consistent Text-to-3D Generation

Qing Zhang , Jinguang Tong , Jing Zhang , Jie Hong , Xuesong Li This is my paper

Pith reviewed 2026-05-20 05:26 UTC · model grok-4.3

classification 💻 cs.CV
keywords text-to-3D generationJanus problemmulti-view consistencydiffusion priorsstructural energyU-Net featuresPCA subspacedenoising guidance
0
0 comments X p. Extension
pith:OYP74K7Q Add to your LaTeX paper What is a Pith Number? 
\usepackage{pith}
\pithnumber{OYP74K7Q}

Prints a linked pith:OYP74K7Q badge after your title and writes the identifier into PDF metadata. Compiles on arXiv with no extra files. Learn more

The pith

Structural energy from U-Net PCA features guides denoising to fix viewpoint inconsistencies in text-to-3D generation.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper argues that the Janus problem in text-to-3D generation, where models display mismatched geometry across viewpoints, arises mainly from viewpoint bias in 2D diffusion priors. It introduces Structural Energy-Guided Sampling, a plug-and-play method that forms a structural energy term inside the PCA subspace of U-Net features and adds the term's gradient to the denoising steps. This runs without any retraining and slots into existing SDS and VSD pipelines. If the approach holds, generators would produce 3D objects whose shapes remain consistent from every angle while textures stay true to the prompt. Experiments across DreamFusion, Magic3D, and LucidDreamer report lower inconsistency rates and higher view-consistency scores.

Core claim

The central claim is that viewpoint bias in 2D diffusion priors produces the Janus problem, and that constructing structural energy in the PCA subspace of U-Net features and injecting its gradient during denoising corrects the bias, yielding more consistent multi-view geometry without harming appearance fidelity.

What carries the argument

Structural Energy-Guided Sampling (SEGS), which extracts principal components from U-Net features, forms a structural energy function on that subspace, and supplies the energy gradient to steer the diffusion trajectory.

If this is right

  • SEGS adds directly to SDS and VSD pipelines without any model retraining or fine-tuning.
  • Average Janus Rate drops by roughly 10 percent across tested baselines.
  • View-CS scores rise, indicating stronger geometric agreement across rendered viewpoints.
  • Appearance fidelity is preserved, so prompt-aligned textures and details remain intact.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Structural constraints in feature subspaces may offset other biases that appear when 2D models are repurposed for 3D tasks.
  • Varying the number of PCA components or the choice of U-Net layer could produce further gains in consistency.
  • The same energy-injection idea might extend to other generative consistency problems such as temporal coherence in video synthesis.

Load-bearing premise

Viewpoint bias in the 2D diffusion prior is the dominant cause of the Janus problem and can be corrected by adding a gradient from structural energy computed in the PCA subspace of U-Net features.

What would settle it

Apply the same set of text prompts to a baseline generator both with and without the structural energy gradient, then measure the Janus Rate on the resulting 3D outputs; absence of a clear reduction would show the guidance does not address the claimed cause.

read the original abstract

Text-to-3D generation based on diffusion models often suffers from the Janus problem, leading to inconsistent geometry across viewpoints. This work identifies viewpoint bias in 2D diffusion priors as the main cause and proposes Structural Energy-Guided Sampling (SEGS), a training-free and plug-and-play framework to improve multi-view consistency. SEGS constructs a structural energy in the PCA subspace of U-Net features and injects its gradient into the denoising process. It can be easily integrated into SDS/VSD pipelines without retraining. Experiments show that SEGS reduces the Janus Rate by about 10% on average and improves View-CS scores across multiple baselines, including DreamFusion, Magic3D, and LucidDreamer. This method effectively alleviates viewpoint artifacts while preserving appearance fidelity, providing a flexible solution for high-quality text-to-3D content generation.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper identifies viewpoint bias in 2D diffusion priors as the main driver of the Janus problem in text-to-3D generation. It proposes Structural Energy-Guided Sampling (SEGS), a training-free plug-and-play method that constructs a structural energy in the PCA subspace of U-Net features and injects the resulting gradient into SDS/VSD denoising. Experiments claim an average ~10% reduction in Janus Rate and gains in View-CS scores when applied to DreamFusion, Magic3D, and LucidDreamer while preserving appearance fidelity.

Significance. If the central claim holds, SEGS offers a lightweight, training-free way to improve multi-view consistency across existing text-to-3D pipelines. The plug-and-play integration without retraining is a practical strength that could see broad adoption for reducing viewpoint artifacts in generated 3D assets.

major comments (2)
  1. [§4] §4 (Experiments): The reported average 10% Janus Rate reduction and View-CS improvements are presented without error bars, number of runs, dataset statistics, or statistical significance tests. This absence makes it impossible to determine whether the gains exceed evaluation variance and directly undermines verification of the central claim.
  2. [§3.2] §3.2 (Structural Energy Construction): The manuscript does not demonstrate that the PCA-derived energy systematically assigns lower values to view-consistent 3D configurations than to Janus configurations. Without this verification or an ablation isolating the PCA subspace from generic feature statistics, the observed improvements could result from incidental regularization rather than targeted correction of viewpoint bias.
minor comments (2)
  1. [Abstract] Abstract: Replace the vague 'about 10%' with the precise average value and standard deviation from the experimental tables.
  2. [§2] §2 (Related Work): The discussion of prior viewpoint-bias mitigation techniques could include more recent references on U-Net feature analysis for geometry.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their insightful comments on our manuscript. We address each of the major comments below and will revise the paper to incorporate additional experimental details and analyses as suggested.

read point-by-point responses

  1. Referee: [§4] §4 (Experiments): The reported average 10% Janus Rate reduction and View-CS improvements are presented without error bars, number of runs, dataset statistics, or statistical significance tests. This absence makes it impossible to determine whether the gains exceed evaluation variance and directly undermines verification of the central claim.

    Authors: We agree with this observation. The current presentation of results does not include sufficient statistical details. In the revised manuscript, we will rerun the experiments with multiple random seeds, report mean and standard deviation for the Janus Rate and View-CS metrics, include the number of runs and dataset statistics, and add statistical significance tests to confirm that the improvements are meaningful beyond variance. revision: yes

  2. Referee: [§3.2] §3.2 (Structural Energy Construction): The manuscript does not demonstrate that the PCA-derived energy systematically assigns lower values to view-consistent 3D configurations than to Janus configurations. Without this verification or an ablation isolating the PCA subspace from generic feature statistics, the observed improvements could result from incidental regularization rather than targeted correction of viewpoint bias.

    Authors: This is a valid point regarding the need for more direct validation of the energy function. To address it, we will include in the revision a new experiment or figure that evaluates the structural energy on both view-consistent and Janus-affected 3D generations. Additionally, we will provide an ablation study comparing the PCA subspace to the full U-Net feature space to demonstrate that the PCA projection is key to targeting the viewpoint bias rather than providing generic regularization. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation relies on external U-Net features

full rationale

The paper constructs structural energy explicitly from the PCA subspace of existing U-Net features extracted during denoising and injects its gradient into SDS/VSD sampling. This step uses pre-trained diffusion model activations as an independent input rather than defining the energy in terms of the target multi-view consistency metric or Janus rate. Empirical gains (10% Janus-rate reduction, View-CS improvement) are reported as measured outcomes across baselines, not as predictions forced by construction. No self-citation chains, uniqueness theorems from prior author work, or fitted parameters renamed as predictions appear in the described method. The approach remains training-free and plug-and-play, keeping the central derivation self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract provides no explicit free parameters, axioms, or invented entities; the approach relies on standard diffusion sampling and PCA without stating new assumptions beyond the identified viewpoint bias.

pith-pipeline@v0.9.0 · 5680 in / 1145 out tokens · 51622 ms · 2026-05-20T05:26:19.852566+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

46 extracted references · 46 canonical work pages · 2 internal anchors

  1. [1]

    Advances in Neural Information Processing Systems35, 25278–25294 (2022)

    Schuhmann, C., Beaumont, R., Vencu, R., Gordon, C., Wightman, R., Cherti, M., Coombes, T., Katta, A., Mullis, C., Wortsman, M.,et al.: Laion-5b: An open large-scale dataset for training next generation image- text models. Advances in Neural Information Processing Systems35, 25278–25294 (2022)

    work page 2022

  2. [2]

    Advances in Neural Information Processing Systems36(2024)

    Deitke, M., Liu, R., Wallingford, M., Ngo, H., Michel, O., Kusupati, A., Fan, A., Laforte, C., Voleti, V., Gadre, S.Y., et al.: Objaverse-xl: A uni- verse of 10m+ 3d objects. Advances in Neural Information Processing Systems36(2024)

    work page 2024

  3. [3]

    Nature medicine31(10), 3404– 3413 (2025)

    Wu, Y., Qian, B., Li, T., Qin, Y., Guan, Z., Chen, T., Jia, Y., Zhang, P., Zeng, D., Moroi, S.,et al.: An eyecare foundation model for clinical assistance: a randomized controlled trial. Nature medicine31(10), 3404– 3413 (2025)

    work page 2025

  4. [4]

    In: The Eleventh International Conference on Learning Representations (ICLR) (2023)

    Poole, B., Jain, A., Barron, J.T., Mildenhall, B.: Dreamfusion: Text-to-3d using 2d diffusion. In: The Eleventh International Conference on Learning Representations (ICLR) (2023)

    work page 2023

  5. [5]

    In: Proceed- ings of the IEEE/CVF International Conference on Computer Vision, pp

    Chen, R., Chen, Y., Jiao, N., Jia, K.: Fantasia3d: Disentangling geometry and appearance for high-quality text-to-3d content creation. In: Proceed- ings of the IEEE/CVF International Conference on Computer Vision, pp. 22246–22256 (2023)

    work page 2023

  6. [6]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Lin, C.-H., Gao, J., Tang, L., Takikawa, T., Zeng, X., Huang, X., Kreis, K., Fidler, S., Liu, M.-Y., Lin, T.-Y.: Magic3d: High-resolution text-to- 3d content creation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 300–309 (2023)

    work page 2023

  7. [7]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Metzer, G., Richardson, E., Patashnik, O., Giryes, R., Cohen-Or, D.: Latent-nerf for shape-guided generation of 3d shapes and textures. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 12663–12673 (2023)

    work page 2023

  8. [8]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Liang, Y., Yang, X., Lin, J., Li, H., Xu, X., Chen, Y.: Luciddreamer: Towards high-fidelity text-to-3d generation via interval score matching. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 6517–6526 (2024)

    work page 2024

  9. [9]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp

    Zhang, Q., Tong, J., Zhang, J., Hong, J., Li, X.: Improving viewpoint consistency in 3d generation via structure feature and clip guidance. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 6440–6449 (2025) 21

    work page 2025

  10. [10]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Han, P., Ye, C., Zhou, J., Zhang, J., Hong, J., Li, X.: Latent-based diffu- sion model for long-tailed recognition. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 2639–2648 (2024)

    work page 2024

  11. [11]

    Advances in Neural Information Processing Systems36 (2024)

    Wang, Z., Lu, C., Wang, Y., Bao, F., Li, C., Su, H., Zhu, J.: Prolific- dreamer: High-fidelity and diverse text-to-3d generation with variational score distillation. Advances in Neural Information Processing Systems36 (2024)

    work page 2024

  12. [12]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Cao, Y., Cao, Y.-P., Han, K., Shan, Y., Wong, K.-Y.K.: Dreamavatar: Text-and-shape guided 3d human avatar generation via diffusion models. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 958–968 (2024)

    work page 2024

  13. [13]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp

    Liu, R., Wu, R., Van Hoorick, B., Tokmakov, P., Zakharov, S., Vondrick, C.: Zero-1-to-3: Zero-shot one image to 3d object. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 9298–9309 (2023)

    work page 2023

  14. [14]

    In: The Twelfth International Conference on Learning Representations (ICLR) (2024)

    Shi, Y., Wang, P., Ye, J., Mai, L., Li, K., Yang, X.: Mvdream: Multi-view diffusion for 3d generation. In: The Twelfth International Conference on Learning Representations (ICLR) (2024)

    work page 2024

  15. [15]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Huang, T., Zeng, Y., Zhang, Z., Xu, W., Xu, H., Xu, S., Lau, R.W., Zuo, W.: Dreamcontrol: Control-based text-to-3d generation with 3d self-prior. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 5364–5373 (2024)

    work page 2024

  16. [16]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Tumanyan, N., Geyer, M., Bagon, S., Dekel, T.: Plug-and-play diffusion features for text-driven image-to-image translation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 1921–1930 (2023)

    work page 1921

  17. [17]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Mo, S., Mu, F., Lin, K.H., Liu, Y., Guan, B., Li, Y., Zhou, B.: Freecontrol: Training-free spatial control of any text-to-image diffusion model with any condition. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 7465–7475 (2024)

    work page 2024

  18. [18]

    Point-E: A System for Generating 3D Point Clouds from Complex Prompts

    Nichol, A., Jun, H., Dhariwal, P., Mishkin, P., Chen, M.: Point-e: A system for generating 3d point clouds from complex prompts. arXiv preprint arXiv:2212.08751 (2022)

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  19. [19]

    Shap-E: Generating Conditional 3D Implicit Functions

    Jun, H., Nichol, A.: Shap-e: Generating conditional 3d implicit functions. arXiv preprint arXiv:2305.02463 (2023)

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  20. [20]

    arXiv preprint arXiv:2402.01166 (2024)

    Liu, J., Huang, X., Huang, T., Chen, L., Hou, Y., Tang, S., Liu, Z., 22 Ouyang, W., Zuo, W., Jiang, J., et al.: A comprehensive survey on 3d content generation. arXiv preprint arXiv:2402.01166 (2024)

    work page arXiv 2024

  21. [21]

    Communications of the ACM65(1), 99–106 (2021)

    Mildenhall, B., Srinivasan, P.P., Tancik, M., Barron, J.T., Ramamoorthi, R., Ng, R.: Nerf: Representing scenes as neural radiance fields for view synthesis. Communications of the ACM65(1), 99–106 (2021)

    work page 2021

  22. [22]

    arXiv preprint arXiv:2501.13104 (2025)

    Xiao, W., Chierchia, R., Cruz, R.S., Li, X., Ahmedt-Aristizabal, D., Sal- vado, O., Fookes, C., Lebrat, L.: Neural radiance fields for the real world: A survey. arXiv preprint arXiv:2501.13104 (2025)

    work page arXiv 2025

  23. [23]

    The Visual Computer 42(1), 118 (2026)

    Dong, Z., Yu, T.: Swiftcraft3d: semantic-enhanced multi-view prompting for efficient and high-fidelity text-to-3d generation. The Visual Computer 42(1), 118 (2026)

    work page 2026

  24. [24]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Yi, T., Fang, J., Wang, J., Wu, G., Xie, L., Zhang, X., Liu, W., Tian, Q., Wang, X.: Gaussiandreamer: Fast generation from text to 3d gaussians by bridging 2d and 3d diffusion models. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 6796–6807 (2024)

    work page 2024

  25. [25]

    In: The Twelfth International Conference on Learning Representations (ICLR) (2024)

    Tang, J., Ren, J., Zhou, H., Liu, Z., Zeng, G.: Dreamgaussian: Genera- tive gaussian splatting for efficient 3d content creation. In: The Twelfth International Conference on Learning Representations (ICLR) (2024)

    work page 2024

  26. [26]

    ACM Trans

    Kerbl, B., Kopanas, G., Leimk¨ uhler, T., Drettakis, G.: 3d gaussian splat- ting for real-time radiance field rendering. ACM Trans. Graph.42(4), 139–1 (2023)

    work page 2023

  27. [27]

    In: Proceedings of the Computer Vision and Pattern Recognition Conference, pp

    Tong, J., Li, X., Maken, F.A., Muthu, S., Petersson, L., Nguyen, C., Li, H.: Gs-2dgs: Geometrically supervised 2dgs for reflective object reconstruc- tion. In: Proceedings of the Computer Vision and Pattern Recognition Conference, pp. 21547–21557 (2025)

    work page 2025

  28. [28]

    In: Proceedings of the 33rd ACM International Conference on Multimedia, pp

    Li, X., Tong, J., Hong, J., Rolland, V., Petersson, L.: Dgns: Deformable gaussian splatting and dynamic neural surface for monocular dynamic 3d reconstruction. In: Proceedings of the 33rd ACM International Conference on Multimedia, pp. 1812–1821 (2025)

    work page 2025

  29. [29]

    The Visual Computer40(7), 4701–4712 (2024)

    Xu, H., Wu, Y., Tang, X., Zhang, J., Zhang, Y., Zhang, Z., Li, C., Jin, X.: Fusiondeformer: text-guided mesh deformation using diffusion models. The Visual Computer40(7), 4701–4712 (2024)

    work page 2024

  30. [30]

    gao et al

    Gao, W., Li, X., Liu, C., Wang, J., Yu, D.: Disentangled text-driven styl- ization of 3d faces via directional clip losses: W. gao et al. The Visual Computer41(12), 10451–10466 (2025) 23

    work page 2025

  31. [31]

    In: The Twelfth International Conference on Learning Representations (ICLR) (2024)

    Li, J., Tan, H., Zhang, K., Xu, Z., Luan, F., Xu, Y., Hong, Y., Sunkavalli, K., Shakhnarovich, G., Bi, S.: Instant3d: Fast text-to-3d with sparse-view generation and large reconstruction model. In: The Twelfth International Conference on Learning Representations (ICLR) (2024)

    work page 2024

  32. [32]

    In: European Conference on Computer Vision, pp

    Tang, J., Chen, Z., Chen, X., Wang, T., Zeng, G., Liu, Z.: Lgm: Large multi-view gaussian model for high-resolution 3d content creation. In: European Conference on Computer Vision, pp. 1–18 (2024). Springer

    work page 2024

  33. [33]

    IEEE Transactions on Pattern Analysis and Machine Intelligence (2025)

    Wang, J., Lu, X., Bennamoun, M., Sheng, B.: Non-rigid point cloud reg- istration via anisotropic hybrid field harmonization. IEEE Transactions on Pattern Analysis and Machine Intelligence (2025)

    work page 2025

  34. [34]

    In: Proceedings of the 31st ACM International Conference on Multimedia, pp

    Chen, Y., Pan, Y., Li, Y., Yao, T., Mei, T.: Control3d: Towards control- lable text-to-3d generation. In: Proceedings of the 31st ACM International Conference on Multimedia, pp. 1148–1156 (2023)

    work page 2023

  35. [35]

    Advances in Neural Information Processing Systems36, 11970–11987 (2023)

    Hong, S., Ahn, D., Kim, S.: Debiasing scores and prompts of 2d diffusion for view-consistent text-to-3d generation. Advances in Neural Information Processing Systems36, 11970–11987 (2023)

    work page 2023

  36. [36]

    arXiv preprint arXiv:2304.04968 (2023)

    Armandpour, M., Sadeghian, A., Zheng, H., Sadeghian, A., Zhou, M.: Re-imagine the negative prompt algorithm: Transform 2d diffusion into 3d, alleviate janus problem and beyond. arXiv preprint arXiv:2304.04968 (2023)

    work page arXiv 2023

  37. [37]

    Advances in neural information processing systems33, 6840–6851 (2020)

    Ho, J., Jain, A., Abbeel, P.: Denoising diffusion probabilistic models. Advances in neural information processing systems33, 6840–6851 (2020)

    work page 2020

  38. [38]

    In: International Conference on Learning Representations (ICLR) (2021)

    Song, J., Meng, C., Ermon, S.: Denoising diffusion implicit models. In: International Conference on Learning Representations (ICLR) (2021)

    work page 2021

  39. [39]

    Advances in neural information processing systems34, 8780–8794 (2021)

    Dhariwal, P., Nichol, A.: Diffusion models beat gans on image synthesis. Advances in neural information processing systems34, 8780–8794 (2021)

    work page 2021

  40. [40]

    Advances in Neural Information Processing Systems36, 16222–16239 (2023)

    Epstein, D., Jabri, A., Poole, B., Efros, A., Holynski, A.: Diffusion self-guidance for controllable image generation. Advances in Neural Information Processing Systems36, 16222–16239 (2023)

    work page 2023

  41. [41]

    In: International Conference on Learning Representations (ICLR) (2021)

    Song, Y., Sohl-Dickstein, J., Kingma, D.P., Kumar, A., Ermon, S., Poole, B.: Score-based generative modeling through stochastic differen- tial equations. In: International Conference on Learning Representations (ICLR) (2021)

    work page 2021

  42. [42]

    Springer, Berlin, Heidelberg (2013) 24

    Oksendal, B.: Stochastic Differential Equations: an Introduction with Applications. Springer, Berlin, Heidelberg (2013) 24

    work page 2013

  43. [43]

    Springer (1996)

    Risken, H.: The Fokker-Planck Equation. Springer (1996)

    work page 1996

  44. [44]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Rombach, R., Blattmann, A., Lorenz, D., Esser, P., Ommer, B.: High- resolution image synthesis with latent diffusion models. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 10684–10695 (2022)

    work page 2022

  45. [45]

    In: International Conference on Machine Learning, pp

    Radford, A., Kim, J.W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J.,et al.: Learning transfer- able visual models from natural language supervision. In: International Conference on Machine Learning, pp. 8748–8763 (2021). PMLR

    work page 2021

  46. [46]

    In: ICCV (2023)

    Liu, Y.-T., Guo, Y.-C., Voleti, V., Shao, R., Chen, C.-H., Luo, G., Zou, Z., Wang, C., Laforte, C., Cao, Y.-P.,et al.: Threestudio: A modular framework for diffusion-guided 3d generation. In: ICCV (2023)

    work page 2023