Pith. sign in

REVIEW 3 major objections 7 minor 48 references

Improving Viewpoint Consistency in 3D Generation via Structure Feature and CLIP Guidance

T0 review · 3 major / 7 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read A tuning-free plugin called ACG cuts the Janus Problem rate in text-to-3D generation from 75–80% to 30–35% without extra optimization time.

desk verdict A plausible plug-and-play fix for the Janus Problem whose headline numbers are not yet statistically established; the method is worth a serious referee but the evaluation needs major work. read the letter →

arxiv 2412.02287 v4 pith:XWLZORMG submitted 2024-12-03 cs.CV

Qing Zhang , Jinguang Tong , Jing Zhang , Jie Hong , Xuesong Li This is my paper · ORCID
classification cs.CV
keywords JanusProblemtext-to-3DgenerationScoreDistillationSamplingcross-attentioncontrolCLIPpruningviewpointconsistencydiffusionmodelbiascoarse-to-finepromptoptimization
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

Text-to-3D generation via Score Distillation Sampling often produces objects with faces on the back, the Janus Problem. This paper argues that the root cause is a viewpoint generation bias in diffusion models: their training data has a long-tailed distribution favoring front views, and complex prompts dilute attention to viewpoint words. The proposed tuning-free mechanism, ACG, counteracts this bias by amplifying viewpoint keywords in cross-attention maps, pruning optimization steps whose pseudo-ground-truth images fail a CLIP similarity check with the requested view, and staging the prompt from object to description. On 20 prompts, adding ACG lowers the Janus Problem Rate from 75–80% in DreamFusion, Magic3D, and LucidDreamer to 30–35%, with no extra optimization time or VRAM overhead. Read sympathetically, the paper establishes viewpoint distribution imbalance inside SDS guidance as the mechanism and shows that a cheap, plug-in rebalancing is sufficient to reduce the artifact.

What carries the argument

The load-bearing object is the cross-attention map $M$ of the U-Net, the per-token weight matrix that connects latent image features to prompt tokens (Eq. 10), together with the CLIP similarity $\sigma$ between the pseudo-ground-truth image and the viewpoint prompt (Eq. 12). The cross-attention map is modified elementwise by $M' = M \odot (1 + \lambda I_{\mathrm{keywords}})$ to boost or suppress viewpoint tokens such as "front", "side", and "back", pushing generation into low-probability viewpoint regions without retraining. The CLIP score is used as a gate: if $\sigma$ falls below $\tau = \alpha \sigma_{\min} + (1-\alpha)\sigma_{\mathrm{mean}}$, where $\sigma_{\min}$ and $\sigma_{\mathrm{mean}}$ are collected during a first unpruned epoch, that optimization step is skipped so wrong-view guidance never enters the SDS gradient. This machinery rebalances the distribution of pseudo-GT viewpoints and, according to the paper, is what lowers the Janus Problem Rate.

What would settle it

Run the same 20-prompt Janus Problem Rate benchmark with CLIP pruning disabled, or with CLIP similarity scores replaced by random values; if JR stays near 30–35%, then CLIP pruning is not the load-bearing component and the claimed mechanism is wrong.

Watch

Extended reading notes

Core claim

The central claim is that the Janus Problem is not an intrinsic failure of 3D optimization but a guidance imbalance inherited from the diffusion prior. Concretely, the paper rewrites SDS as a difference between the rendered image and a pseudo-ground-truth $\bar{x}_0$ from the denoiser (Eq. 8–9), and argues that because the diffusion model's training data over-represents front views, the pseudo-GT distribution is front-heavy, driving the NeRF to grow extra faces at every angle. Two supporting observations are made: a sample of 6,762 images from LAION-5B shows a long-tailed viewpoint distribution, and cross-attention maps show attention to the token "back" dropping as prompt complexity increases. ACG then acts in three places: $M' = M \odot (1 + \lambda I_{\mathrm{keywords}})$ raises attention to viewpoint words, a CLIP similarity score $\sigma$ with an adaptive threshold $\tau = \alpha \sigma_{\min} + (1-\alpha)\sigma_{\mathrm{mean}}$ skips optimizations whose pseudo-GT does not match the desired view, and a two-stage coarse-to-fine prompt schedule protects geometry before details are added. The measured consequence is that the pseudo-GT viewpoint ratio moves from about 2:1:1 to 1.2:1:1 across front/side/back, and the Janus Problem Rate drops to 30–35% on all three baselines.

Load-bearing premise

The whole fix rides on one premise: CLIP can reliably tell, from a rough partially denoised image, whether the viewpoint is wrong, so that dropping those updates removes only bad advice; the authors concede that CLIP was not made for viewpoint recognition, and the threshold's $\alpha$ value is not reported.

Editorial extensions

If this is right

  • ACG is a drop-in for SDS-based text-to-3D pipelines: it alters only cross-attention, filtering, and prompt staging, so DreamFusion, Magic3D, and LucidDreamer each improved without changing their diffusion weights.
  • A direct mechanism is reported: CLIP Pruning shifts the pseudo-GT viewpoint distribution from roughly 2:1:1 to 1.2:1:1 (front:side:back), so the 3D model receives more balanced guidance at back views.
  • Because no fine-tuning is involved, there is no catastrophic-forgetting or per-method retraining cost, and generation speed is unaffected.
  • The paper also reports a boundary: extremely rare object categories still fail, and CLIP gives only coarse viewpoint guidance, so ACG is a mitigation rather than a complete cure.
  • If the reported numbers hold, the same recipe could be applied to other long-tailed attributes in 2D priors, such as pose or lighting, whenever a similarity score can gate the guidance.

Reading between the lines

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

  • A quantitative ablation is missing from the paper: Table 1 reports Janus Problem Rate for full ACG only, while the module studies are qualitative. Rerunning the 20-prompt benchmark with each module disabled would reveal which component actually carries the gain.
  • The threshold formula Eq. 13 contains a hyperparameter $\alpha$, but the paper does not report the value used in the experiments; sweeping $\alpha$ and reporting JR as a function of it would test robustness and enable reproduction.
  • Because CLIP was trained on web images, its viewpoint judgments may encode some of the same front-view bias the paper is trying to remove; fine-tuning CLIP on noise-corrupted renderings, which the paper lists as future work, could sharpen pruning and lower JR further.
  • The Fokker-Planck derivation motivates the root-cause story, but the falsifiable content is the distribution shift and the JR numbers; a broader evaluation across more prompts, random seeds, and object categories would show whether the 1.2:1:1 rebalancing generalizes.
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

3 major / 7 minor

Summary. The paper argues that the Janus Problem in text-to-3D generation stems from viewpoint generation bias in diffusion models, caused by the long-tailed viewpoint distribution of training data and by reduced attention to viewpoint tokens as prompts become more complex. To counter this, the authors propose ACG, a tuning-free plug-in that (i) amplifies cross-attention weights on viewpoint keywords, (ii) prunes SDS updates whose CLIP similarity to the expected view falls below an adaptive threshold, and (iii) uses a coarse-to-fine prompt splitting strategy. Experiments on DreamFusion, Magic3D, and LucidDreamer report a reduction in Janus Problem Rate from 75–80% to 30–35%, together with qualitative and ablation evidence that each component contributes to the improvement.

Significance. If the central claims are established, ACG would be a practically useful contribution: it is tuning-free, compatible with several existing text-to-3D frameworks, and does not add optimization time or VRAM overhead according to the qualitative comparison in Figure 6. The paper also provides a useful empirical observation—that viewpoint attention in cross-attention maps weakens as prompts become more complex (Figure 3)—and demonstrates a concrete mechanism for rebalancing pseudo-GT viewpoint distributions (Figure 10). However, the load-bearing quantitative evidence, especially the Janus Problem Rate in Table 1, is currently reported without the statistical detail needed to support the headline 40–50 percentage-point improvement, and the root-cause derivation in Section 3.1 does not fully close the gap between standard SDE theory and the specific claim of long-tailed viewpoint bias. The paper's own Limitation paragraph concedes that CLIP is not designed for viewpoint recognition, which further tempers the strength of the pruning claim. With strengthened evaluation and clearer specification of hyperparameters, the contribution would be a solid incremental advance.

major comments (3)
  1. [§3.1, Eqs. (5)–(9)] The derivation from the Fokker–Planck equations and the re-expression of SDS does not establish that the long-tailed viewpoint distribution of training data creates biased pseudo-GT guidance. Equations (5) and (6) are standard results about the evolution of the data distribution under the forward and reverse SDEs; they hold regardless of the marginal viewpoint distribution. Equation (9) restates SDS as a difference between the rendered image and the predicted pseudo-GT, but nothing in these equations shows that a long-tailed viewpoint marginal in p0(x) produces a score bias at the specific render angles used in 3D optimization. The claim is plausible and supported qualitatively by Figure 2, but the paper presents it as a 'demonstration' rather than as an empirical hypothesis. Please either provide a concrete mathematical link between the viewpoint marginal and the SDS gradient, or soften the language to present the long-tailed bias as a motivated empirical observation.
  2. [§4.1, Table 1] The central quantitative claim—that ACG reduces JR from 75–80% to 30–35%—is not statistically supported as reported. The text states only that 20 prompts were randomly selected and that inconsistent content was counted, with no definition of 'inconsistent', no raw counts per prompt, no seeds or repeated runs, no confidence intervals, and no inter-rater reliability. With n=20, the standard error of a 35% proportion is roughly 10 percentage points, so the reported 45-point drop is statistically indistinguishable from a much smaller effect under a simple binomial model. Additionally, the comparison does not isolate the proposed viewpoint-selective pruning from the generic effect of skipping optimization updates. Since CLIP Pruning drops SDS steps below a threshold, a control that skips the same number of updates uniformly at random is needed to rule out the possibility that any reduction in guidance frequency improves consistency. I recommend reporting the full per-prompt counts, a scoring rubric, significance testing, and the random-skip control.
  3. [§3.3, Eq. (13)] The adaptive threshold for CLIP Pruning depends on an unreported hyperparameter α, and its operational assumption is not validated. Equation (13) defines τ = α·σ_min + (1−α)·σ_mean, but the paper never reports the value of α used in the experiments, nor its sensitivity. Since α directly controls how many updates are pruned, the reported JR improvements cannot be reproduced without this value. Moreover, the assumption that low CLIP similarity between a noisy pseudo-GT image and the expected view text reliably identifies viewpoint mismatch is not tested; the authors' own Limitation paragraph concedes that CLIP is not built for viewpoint recognition and provides only coarse guidance. I ask for the α value, an ablation over α, and a validation that the pruned pseudo-GTs are indeed viewpoint-mismatched rather than merely low-quality or noisy.
minor comments (7)
  1. [§3.1, Eq. (7)] Equation (7) uses ϵ both for the standard Gaussian noise and for the rendered image x0 = g(θ,v), which is confusing; please introduce a distinct symbol for the rendered image, e.g., x0 = g(θ,v), and keep ϵ for noise.
  2. [§3.1, Eq. (9)] The transformation from Equation (7) to Equation (9) assumes a specific relationship between the noise scaling and the weighting ω(t); please state the assumption explicitly so that the algebraic step is verifiable.
  3. [§3.1] The phrase 'This chapter aims to demonstrate' should be 'This section aims to demonstrate' for consistency with the paper's structure.
  4. [§4.1] The text says 'the multi-head problem' in the qualitative comparison; this appears to be a typo for 'the Janus problem.'
  5. [Figure 2 caption] The caption contains a placeholder citation '[? ]' for CLIP ViT-H/14; please insert the proper reference.
  6. [§3.2, Eq. (11)] After amplifying the cross-attention map with M' = M ⊙ (1 + λ·I_keywords), the attention weights are no longer normalized; please state whether re-normalization is applied and, if not, explain why the unnormalized map remains a valid attention map.
  7. [Figure 6 caption] The claim that ACG does so 'without extra optimisation time or VRAM overhead' is not backed by any runtime or memory measurements in the paper; please either add a table with wall-clock times and VRAM usage or soften the claim.

Circularity Check

1 steps flagged · score 2.0 of 10

Minor self-referential CLIP use in diagnosis and pruning; no circularity in the headline JR claim.

  1. other [Section 3.3 (CLIP Pruning) and Section 4.2 (Viewpoint Rebalancing, Figure 10)]
    "When the rendering viewpoint falls within this range, if the pseudo-GT image from the Diffusion Model is identified as front-facing, that optimization is pruned... After applying our pruning technique, we observed a significant improvement in balance, achieving a more equitable distribution of 1.2:1:1."

    The pruning rule removes exactly the pseudo-GT images that the CLIP-based identification labels as mismatched to the desired viewpoint, and the post-pruning viewpoint distribution in Figure 10 is then assessed with the same kind of CLIP-based viewpoint identification. The reported shift from an approximate 2:1:1 front:side:back ratio to 1.2:1:1 is therefore a direct consequence of the selection rule rather than an independent confirmation that the Janus geometry error has been reduced. This auxiliary demonstration is self-consistent by construction.

full rationale

The paper's central quantitative claim is that ACG lowers the Janus Problem Rate from 75–80% to 30–35% across DreamFusion, Magic3D, and LucidDreamer. That score is obtained from human-inspected renderings of 3D objects (Section 4.1), not from the paper's own viewpoint-bias estimator, so the headline result has independent grounding. The self-referential element is limited to the use of CLIP both to demonstrate the long-tailed viewpoint distribution in Figure 2 and to implement CLIP Pruning in Section 3.3; a shared viewpoint bias in the CLIP model family could influence both the diagnosis and the remedy. Additionally, Figure 10's post-pruning viewpoint rebalancing is partly by construction because the pruning rule removes exactly the samples the same style of CLIP identification labels as mismatched, making that particular demonstration a consistency check rather than independent evidence. No equation in the derivation reduces to its inputs: the Fokker-Planck analysis is standard diffusion mathematics used for motivation, the SDS reinterpretation is algebraic, and the cross-attention control formula is an explicit design choice. Self-citations appear in related-work and long-tailed-data contexts but are not load-bearing for the Janus reduction claim. The unreported alpha in the adaptive threshold and the informal JR counting procedure are reporting weaknesses, not circularity.

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

The paper introduces no new physical or conceptual entities; it only combines existing mechanisms. Its empirical claims rest on hand-set hyperparameters (lambda, alpha) and on the domain assumptions that CLIP can judge viewpoints and that attention amplification shifts the generated view distribution.

free parameters (3)
  • lambda (cross-attention amplification) = scaled by prompt length times 10
    Controls strength of viewpoint-token enhancement in Eq. 11; the exact scaling formula is not given.
  • alpha (CLIP threshold blending) = not specified
    Blends sigma_min and sigma_mean in Eq. 13 to set the pruning threshold; the paper never reports the chosen value.
  • prompt split ratio and stage duration = approximately half the optimization per stage
    The division of prompts and the fraction of iterations per stage are chosen by hand.
assumptions (3)
  • domain assumption Diffusion models trained on long-tailed viewpoint distributions generate samples biased toward front views.
    Section 3.1 claims this as the root cause of the Janus Problem, supported by Figure 2 (a CLIP-based histogram) rather than a mathematical proof.
  • domain assumption CLIP similarity between a noisy pseudo-GT image and the view text is a valid indicator of whether the generated view is correct.
    Section 3.3 relies on Eq. 12 as a pruning signal; the limitation section concedes CLIP is not designed for viewpoint recognition.
  • ad hoc to paper Amplifying cross-attention weights on viewpoint tokens increases the probability of the diffusion model generating that viewpoint.
    Eq. 11 is the mechanism; the paper provides qualitative ablation (Figure 7) but no quantitative link between attention weight and view distribution.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Improving Viewpoint Consistency in 3D Generation via Structure Feature and CLIP Guidance." pith.science (2026). https://pith.science/paper/XWLZORMG

@misc{pith2026241202287,
  author       = {Pith},
  title        = {Pith review of: Improving Viewpoint Consistency in 3D Generation via Structure Feature and CLIP Guidance},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XWLZORMG}},
  note         = {Machine review of arXiv:2412.02287}
}
read the original abstract

Despite recent advances in text-to-3D generation techniques, current methods often suffer from geometric inconsistencies, commonly referred to as the Janus Problem. This paper identifies the root cause of the Janus Problem: viewpoint generation bias in diffusion models, which creates a significant gap between the actual generated viewpoint and the expected one required for optimizing the 3D model. To address this issue, we propose a tuning-free approach called the Attention and CLIP Guidance (ACG) mechanism. ACG enhances desired viewpoints by adaptively controlling cross-attention maps, employs CLIP-based view-text similarities to filter out erroneous viewpoints, and uses a coarse-to-fine optimization strategy with staged prompts to progressively refine 3D generation. Extensive experiments demonstrate that our method significantly reduces the Janus Problem without compromising generation speed, establishing ACG as an efficient, plug-and-play component for existing text-to-3D frameworks.

Figures

Figures reproduced from arXiv: 2412.02287 by the authors.

Figure 1
Figure 1. Viewpoint bias in Stable Diffusion models. We show generation samples from Stable Diffusion v1.4 and v2.1 from the prompt “a corgi puppy”. Most samples show a corgi in a forward￾facing pose. models to optimize Neural Radiance Fields (NeRF) [27], enabling the generation of detailed 3D representations from textual descriptions. Central to this framework is the Score Distillation Sampling (SDS) algorithm, which has dem… view at source ↗
Figure 2
Figure 2. The long-tailed distribution of viewpoints in the diffusion model’s training data biases the generated viewpoints. The left figure shows a random sample of 6,762 images from Laion2B [34], categorized into three object viewpoints using CLIP ViT-H/14 [? ]. The diffusion model maps this long-tailed distribution to a Gaussian distribution, which then produces a similarly biased generation distribution via reverse SDE—fu… view at source ↗
Figure 3
Figure 3. Average attention maps. As the prompt complexity increases, the diffusion model’s attention to viewpoint tokens ”back” decreases significantly. surface optimization, and DreamGaussian [42] employs ex￾plicit representations to accelerate generation by exploiting the fast rendering speed of Gaussian Splatting [18, 20, 43]. The Janus problem, which concerns generating consistent front and back views, is the primary foc… view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Overview of our work. We adjust cross-attention maps in the diffusion model to enhance focus on viewpoint keywords, guiding the model to generate diverse viewpoints. CLIP guidance prunes incorrect viewpoints during denoising based on a loss threshold, ensuring alignmen…
Figure 5
Figure 5. Figure 5: Examples generated by ACG. ACG can enhance the consistency of geometric shapes and the variety of high-fidelity textures in text-to-3D content generation. into a query matrix Q = lQ(ϕ(zt)), while the text prompt P is embedded as ψ(P) and mapped to key and value matrice…
Figure 6
Figure 6. Figure 6: Qualitative comparison with baseline methods in text-to-3D generation. From top to bottom: Stable-DreamFusion, Magic3D, LucidDreamer, and our ACG-enhanced versions. For each prompt, we show the 0°, 90°, 180,° and 270° views. ACG consistently recovers the missing back g…
Figure 7
Figure 7. Figure 7: Ablation study on controlling cross attention map. Controlling Cross Attention can suppress the generation of front views when the rendering viewpoint is at the back view [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Ablation study on coarse to fine optimization based on prompt. For Dreamfusion [41] with seed=2, the top row uses the original prompt “a DSLR photo of a corgi puppy”, and the bottom row only uses “a corgi puppy” [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: Ablation study on CLIP Pruning. These pseudo￾Ground-Truth images identified by CLIP as front-facing are pruned in the back viewpoint range of [120, 240] degrees. This ensures only back-view images are retained, enhancing viewpoint consistency in the generated 3D repres…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 18 canonical work pages

  1. [1]

    Re-imagine the negative prompt algorithm: Transform 2d diffusion into 3d, alleviate janus problem and beyond

    Mohammadreza Armandpour, Ali Sadeghian, Huangjie Zheng, Amir Sadeghian, and Mingyuan Zhou. Re-imagine the negative prompt algorithm: Transform 2d diffusion into 3d, alleviate janus problem and beyond. arXiv preprint arXiv:2304.04968, 2023. 1, 3, 4

    arXiv 2023

  2. [2]

    In- structpix2pix: Learning to follow image editing instructions

    Tim Brooks, Aleksander Holynski, and Alexei A Efros. In- structpix2pix: Learning to follow image editing instructions. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 18392–18402, 2023. 5

    2023

  3. [3]

    Dreamavatar: Text-and-shape guided 3d hu- man avatar generation via diffusion models

    Yukang Cao, Yan-Pei Cao, Kai Han, Ying Shan, and Kwan- Yee K Wong. Dreamavatar: Text-and-shape guided 3d hu- man avatar generation via diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages 958–968, 2024. 1

    work page 2024

  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. 1, 2

    work page 2023

  5. [5]

    Text-to-3d using gaussian splatting

    Zilong Chen, Feng Wang, Yikai Wang, and Huaping Liu. Text-to-3d using gaussian splatting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 21401–21412, 2024. 2

    2024

  6. [6]

    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. 1

    2024

  7. [7]

    Vfusion3d: Learning scalable 3d generative models from video diffusion models

    Junlin Han, Filippos Kokkinos, and Philip Torr. Vfusion3d: Learning scalable 3d generative models from video diffusion models. European Conference on Computer Vision (ECCV),

  8. [8]

    Latent-based diffusion model for long-tailed recognition

    Pengxiao Han, Changkun Ye, Jieming Zhou, Jing Zhang, Jie Hong, and Xuesong Li. Latent-based diffusion model for long-tailed recognition. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 2639–2648, 2024. 2

    work page 2024

Show all 48 references
  1. [9]

    Enhancing features in long-tailed data using large vision model

    Pengxiao Han, Changkun Ye, Jinguang Tong, Cuicui Jiang, Jie Hong, Li Fang, and Xuesong Li. Enhancing features in long-tailed data using large vision model. arXiv preprint arXiv:2504.10852, 2025. 4

    2025

  2. [10]

    Prompt-to-prompt im- age editing with cross attention control

    Amir Hertz, Ron Mokady, Jay Tenenbaum, Kfir Aberman, Yael Pritch, and Daniel Cohen-Or. Prompt-to-prompt im- age editing with cross attention control. arXiv preprint arXiv:2208.01626, 2022. 5

    2022 arXiv

  3. [11]

    Denoising dif- fusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising dif- fusion probabilistic models. Advances in neural information processing systems, 33:6840–6851, 2020. 3, 4

    2020

  4. [12]

    Pointcam: Cut- and-mix for open-set point cloud learning

    Jie Hong, Shi Qiu, Weihao Li, Saeed Anwar, Mehrtash Ha- randi, Nick Barnes, and Lars Petersson. Pointcam: Cut- and-mix for open-set point cloud learning. arXiv preprint arXiv:2212.02011, 2023. 4

    2023 arXiv

  5. [13]

    Debi- asing scores and prompts of 2d diffusion for view-consistent text-to-3d generation

    Susung Hong, Donghoon Ahn, and Seungryong Kim. Debi- asing scores and prompts of 2d diffusion for view-consistent text-to-3d generation. Advances in Neural Information Pro- cessing Systems, 36:11970–11987, 2023. 1, 3

    2023

  6. [14]

    Lora: Low-rank adaptation of large language models

    Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen- Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. Lora: Low-rank adaptation of large language models. arXiv preprint arXiv:2106.09685, 2021. 1

    2021 arXiv

  7. [15]

    Dreamcontrol: Control-based text-to-3d generation with 3d self-prior

    Tianyu Huang, Yihan Zeng, Zhilu Zhang, Wan Xu, Hang Xu, Songcen Xu, Rynson WH Lau, and Wangmeng Zuo. Dreamcontrol: Control-based text-to-3d generation with 3d self-prior. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 5364– 5373, 2...

    2024

  8. [16]

    Zero-shot text-guided object genera- tion with dream fields

    Ajay Jain, Ben Mildenhall, Jonathan T Barron, Pieter Abbeel, and Ben Poole. Zero-shot text-guided object genera- tion with dream fields. In Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition , pages 867–876, 2022. 2

    2022

  9. [17]

    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

  10. [18]

    3d gaussian splatting for real-time radiance field rendering

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk ¨uhler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering. ACM Trans. Graph., 42(4):139–1,

  11. [19]

    Auto-encoding variational bayes

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

    2013 arXiv

  12. [20]

    Dgns: Deformable gaussian splatting and dynamic neural surface for monocular dynamic 3d recon- struction

    Xuesong Li, Jinguang Tong, Jie Hong, Vivien Rolland, and Lars Petersson. Dgns: Deformable gaussian splatting and dynamic neural surface for monocular dynamic 3d recon- struction. arXiv preprint arXiv:2412.03910, 2024. 3

    2024 arXiv

  13. [21]

    Luciddreamer: Towards high- fidelity text-to-3d generation via interval score matching

    Yixun Liang, Xin Yang, Jiantao Lin, Haodong Li, Xiao- gang Xu, and Yingcong Chen. Luciddreamer: Towards high- fidelity text-to-3d generation via interval score matching. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 6517–6526, 20...

    2024

  14. [22]

    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

  15. [23]

    A comprehensive survey on 3d con- tent generation

    Jian Liu, Xiaoshui Huang, Tianyu Huang, Lu Chen, Yuenan Hou, Shixiang Tang, Ziwei Liu, Wanli Ouyang, Wangmeng Zuo, Junjun Jiang, et al. A comprehensive survey on 3d con- tent generation. arXiv preprint arXiv:2402.01166, 2024. 2

    2024 arXiv

  16. [24]

    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. 1, 3

    2023

  17. [25]

    Threestudio: A modular framework for diffusion-guided 3d generation

    Ying-Tian Liu, Yuan-Chen Guo, Vikram V oleti, Ruizhi Shao, Chia-Hao Chen, Guan Luo, Zixin Zou, Chen Wang, Chris- tian Laforte, Yan-Pei Cao, et al. Threestudio: A modular framework for diffusion-guided 3d generation. In ICCV,

  18. [26]

    Latent-nerf for shape-guided generation of 3d shapes and textures

    Gal Metzer, Elad Richardson, Or Patashnik, Raja Giryes, and Daniel Cohen-Or. Latent-nerf for shape-guided generation of 3d shapes and textures. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 12663–12673, 2023. 1

    2023

  19. [27]

    Nerf: Representing scenes as neural radiance fields for view syn- thesis

    Ben Mildenhall, Pratul P Srinivasan, Matthew Tancik, Jonathan T Barron, Ravi Ramamoorthi, and Ren Ng. Nerf: Representing scenes as neural radiance fields for view syn- thesis. Communications of the ACM, 65(1):99–106, 2021. 1, 2

    2021

  20. [28]

    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

    2022 arXiv

  21. [29]

    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. 1, 2, 4, 7

    2022 arXiv

  22. [30]

    Learning transferable visual models from natural language supervi- sion

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervi- sion. In International conference on machine learning, ...

    2021

  23. [31]

    High-resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj ¨orn Ommer. 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. 3, 4

    2022

  24. [32]

    U- net: Convolutional networks for biomedical image segmen- tation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U- net: Convolutional networks for biomedical image segmen- tation. In Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, proceedings, pa...

    2015

  25. [33]

    Dreambooth: Fine tuning text-to-image diffusion models for subject-driven generation

    Nataniel Ruiz, Yuanzhen Li, Varun Jampani, Yael Pritch, Michael Rubinstein, and Kfir Aberman. Dreambooth: Fine tuning text-to-image diffusion models for subject-driven generation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 2250...

    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]

    Generating high fidelity data from low-density regions using diffusion models

    Vikash Sehwag, Caner Hazirbas, Albert Gordo, Firat Oz- genel, and Cristian Canton. Generating high fidelity data from low-density regions using diffusion models. In Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 11492–11501, 2022. 4

    2022

  28. [36]

    Deep marching tetrahedra: a hybrid repre- sentation for high-resolution 3d shape synthesis

    Tianchang Shen, Jun Gao, Kangxue Yin, Ming-Yu Liu, and Sanja Fidler. Deep marching tetrahedra: a hybrid repre- sentation for high-resolution 3d shape synthesis. Advances in Neural Information Processing Systems , 34:6087–6101,

  29. [37]

    Mvdream: Multi-view diffusion for 3d gen- eration

    Yichun Shi, Peng Wang, Jianglong Ye, Mai Long, Kejie Li, and Xiao Yang. Mvdream: Multi-view diffusion for 3d gen- eration. arXiv preprint arXiv:2308.16512, 2023. 1, 3

    2023 arXiv

  30. [38]

    Denoising diffusion implicit models

    Jiaming Song, Chenlin Meng, and Stefano Ermon. Denoising diffusion implicit models. arXiv preprint arXiv:2010.02502, 2020. 4

    2010 arXiv

  31. [39]

    Generative modeling by esti- mating gradients of the data distribution

    Yang Song and Stefano Ermon. Generative modeling by esti- mating gradients of the data distribution. Advances in neural information processing systems, 32, 2019. 3

    2019

  32. [40]

    Score-based generative modeling through stochastic differential equa- tions

    Yang Song, Jascha Sohl-Dickstein, Diederik P Kingma, Ab- hishek Kumar, Stefano Ermon, and Ben Poole. Score-based generative modeling through stochastic differential equa- tions. arXiv preprint arXiv:2011.13456, 2020. 3

    2011 arXiv

  33. [41]

    Stable-dreamfusion: Text-to-3d with stable-diffusion, 2022

    Jiaxiang Tang. Stable-dreamfusion: Text-to-3d with stable-diffusion, 2022. https://github.com/ashawkey/stable- dreamfusion. 7, 8

    2022

  34. [42]

    Dreamgaussian: Generative gaussian splatting for effi- cient 3d content creation

    Jiaxiang Tang, Jiawei Ren, Hang Zhou, Ziwei Liu, and Gang Zeng. Dreamgaussian: Generative gaussian splatting for effi- cient 3d content creation. arXiv preprint arXiv:2309.16653,

    arXiv

  35. [43]

    Gs-2dgs: Geometrically supervised 2dgs for reflective object reconstruction

    Jinguang Tong, Xuesong Li, Fahira Afzal Maken, Sundaram Muthu, Lars Petersson, Chuong Nguyen, and Hongdong Li. Gs-2dgs: Geometrically supervised 2dgs for reflective object reconstruction. In Proceedings of the Computer Vision and Pattern Recognition Conference, pages 21547–215...

    2025

  36. [44]

    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. 1, 2, 4

    2024

  37. [45]

    Neural radiance fields for the real world: A survey

    Wenhui Xiao, Remi Chierchia, Rodrigo Santa Cruz, Xuesong Li, David Ahmedt-Aristizabal, Olivier Salvado, Clinton Fookes, and Leo Lebrat. Neural radiance fields for the real world: A survey. arXiv preprint arXiv:2501.13104,

    arXiv

  38. [46]

    Gaussiandreamer: Fast generation from text to 3d gaussians by bridging 2d and 3d diffusion models

    Taoran Yi, Jiemin Fang, Junjie Wang, Guanjun Wu, Lingxi Xie, Xiaopeng Zhang, Wenyu Liu, Qi Tian, and Xinggang Wang. Gaussiandreamer: Fast generation from text to 3d gaussians by bridging 2d and 3d diffusion models. In Pro- ceedings of the IEEE/CVF Conference on Computer Vision...

    2024

  39. [47]

    Adding conditional control to text-to-image diffusion models

    Lvmin Zhang, Anyi Rao, and Maneesh Agrawala. Adding conditional control to text-to-image diffusion models. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 3836–3847, 2023. 1

    2023

  40. [48]

    Deep long-tailed learning: A survey

    Yifan Zhang, Bingyi Kang, Bryan Hooi, Shuicheng Yan, and Jiashi Feng. Deep long-tailed learning: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(9):10795–10816, 2023. 2

    2023

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.