arxiv: 2605.00345 · v1 · submitted 2026-05-01 · 💻 cs.CV

Recognition: unknown

Pose-Aware Diffusion for 3D Generation

Zihan Zhou , Luxi Chen , Jingzhi Zhou , Yuhao Wan , Min Zhao , Baoyu Fan , Chongxuan Li

Authors on Pith no claims yet

Pith reviewed 2026-05-09 20:10 UTC · model grok-4.3

classification 💻 cs.CV

keywords 3D generationdiffusion modelspose alignmentimage-to-3Dmonocular depthgeometric anchorscene composition

0 comments

The pith

Pose-Aware Diffusion generates 3D geometry directly in observation space by anchoring diffusion to unprojected monocular depth.

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

The paper introduces Pose-Aware Diffusion as an end-to-end framework that synthesizes 3D objects directly in the input view instead of generating in a canonical frame and then applying rotation. It converts monocular depth into a partial point cloud and injects this as an explicit 3D geometric anchor to guide the diffusion process. This native alignment avoids the spatial mismatches that arise when pose is handled separately. A sympathetic reader cares because the result is higher-fidelity 3D assets whose geometry and placement match the original image, enabling more reliable use in downstream tasks such as scene assembly.

Core claim

PAD is an end-to-end diffusion framework that synthesizes 3D geometry directly within the observation space. By unprojecting monocular depth into a partial point cloud and explicitly injecting it as a 3D geometric anchor, PAD abandons canonical assumptions to enforce rigorous spatial supervision. This native generation intrinsically resolves pose ambiguity, producing high-fidelity pose-aligned assets with superior geometric alignment and image-to-3D correspondence.

What carries the argument

Unprojected monocular depth converted to a partial point cloud and injected as a geometric anchor that supplies 3D spatial supervision during diffusion in the observation space.

If this is right

Produces 3D objects with higher geometric alignment to the input image pose than decoupled canonical-then-rotate pipelines.
Yields improved image-to-3D correspondence compared with prior state-of-the-art methods.
Enables compositional 3D scene reconstruction by simple union of independently generated objects while preserving precise spatial layouts.

Where Pith is reading between the lines

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

The same anchor mechanism could be tested with depth from stereo or RGB-D sensors to reduce reliance on monocular depth networks.
The observation-space generation approach may simplify training pipelines by removing the need to learn separate pose normalization steps.
Independent object generation followed by union could support incremental scene building where new objects are added without regenerating the entire scene.

Load-bearing premise

Unprojecting monocular depth into a partial point cloud supplies a sufficiently accurate and complete 3D geometric anchor without introducing new transformation ambiguities or depth estimation errors.

What would settle it

Quantitative comparison of 3D geometric alignment metrics on a test set containing images with deliberately inaccurate depth estimates, checking whether PAD still outperforms baselines that use canonical generation.

Figures

Figures reproduced from arXiv: 2605.00345 by Baoyu Fan, Chongxuan Li, Jingzhi Zhou, Luxi Chen, Min Zhao, Yuhao Wan, Zihan Zhou.

**Figure 2.** Figure 2: Overview of PAD. (a) Existing decoupled paradigms typically follow a canonical-then-rotate approach, generating an object in a canonical space, followed by pose alignment. (b) In contrast, PAD synthesizes complete 3D meshes directly in the observation space. The model concatenates a geometric latent encoded from the visible partial condition point cloud. From a single image, PAD yields a pose-aligned 3D ob… view at source ↗

**Figure 3.** Figure 3: PAD generates in the observation space to overcome the limitations [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

**Figure 4.** Figure 4: Qualitative comparison on posed object generation. [PITH_FULL_IMAGE:figures/full_fig_p012_4.png] view at source ↗

**Figure 5.** Figure 5: Qualitative comparison on compositional scene generation. [PITH_FULL_IMAGE:figures/full_fig_p013_5.png] view at source ↗

**Figure 6.** Figure 6: Ablation study. We evaluate the key components of PAD. Removing the 3D condition (a) or the point-based representation (b) leads to severe structural distortions and spatial misalignments. Omitting depth noise augmentation (c) causes collapsed surfaces and artifacts when handling imperfect depth inputs. In contrast, our full model (d) robustly generates high-fidelity, perfectly aligned geometry. Red boxes … view at source ↗

**Figure 7.** Figure 7: More results of the generated objects [PITH_FULL_IMAGE:figures/full_fig_p021_7.png] view at source ↗

**Figure 8.** Figure 8: More results of the generated compositional scenes [PITH_FULL_IMAGE:figures/full_fig_p022_8.png] view at source ↗

read the original abstract

Generating pose-aligned 3D objects is challenging due to the spatial mismatches and transformation ambiguities inherent in decoupled canonical-then-rotate paradigms. To this end, we introduce Pose-Aware Diffusion (PAD), a novel end-to-end diffusion framework that synthesizes 3D geometry directly within the observation space. By unprojecting monocular depth into a partial point cloud and explicitly injecting it as a 3D geometric anchor, PAD abandons canonical assumptions to enforce rigorous spatial supervision. This native generation intrinsically resolves pose ambiguity, producing high-fidelity pose-aligned assets. Extensive experiments demonstrate that PAD achieves superior geometric alignment and image-to-3D correspondence compared to state-of-the-art methods. Additionally, PAD naturally extends to compositional 3D scene reconstruction via a simple union of independently generated objects, highlighting its robust ability to preserve precise spatial layouts.

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.

Desk Editor's Note private letter to a colleague

PAD anchors diffusion to unprojected monocular depth in observation space to avoid canonical pose issues, but the depth errors could limit how much real improvement it delivers.

read the letter

The core idea here is generating 3D directly in the input view's space rather than a canonical frame. They unproject monocular depth to a partial point cloud and inject it as an explicit anchor inside the diffusion process. This is meant to cut down on transformation ambiguities and make pose alignment more natural from the start. The scene composition trick—union of separate objects—follows directly from that choice and looks practical for downstream use in graphics or robotics pipelines. That part is a clean architectural move and addresses a real friction point in current image-to-3D work. What they do well is keep the pipeline end-to-end and avoid the usual decoupled rotate-after-generate steps. The abstract positions this as delivering better geometric alignment and correspondence than prior methods. The stress-test concern about monocular depth quality is worth watching. Depth estimators bring scale drift and surface errors, and if the model mostly reproduces the anchor instead of correcting it, the reported gains could be overstated. The abstract does not include metrics, ablations on depth noise, or comparisons against ground-truth depth, so it is hard to judge how much the pose-aware mechanism itself contributes versus the quality of the input anchor. No load-bearing equations appear in the summary, so this reads as an engineering choice rather than a new derivation. Citation coverage is not shown in detail, but the framing suggests they engage with recent depth-conditioned diffusion lines. This is aimed at people building controllable 3D generators who need reliable pose consistency. A reader already working on observation-space methods or compositional scenes could extract useful implementation details if the full experiments hold up. I would send it to peer review so referees can check the numbers and ablations directly.

Referee Report

2 major / 2 minor

Summary. The paper proposes Pose-Aware Diffusion (PAD), an end-to-end diffusion model that generates 3D geometry directly in observation space rather than a canonical frame. It unprojects monocular depth to a partial point cloud, injects this as an explicit 3D geometric anchor for spatial supervision, and claims this resolves pose ambiguity to yield superior geometric alignment and image-to-3D correspondence versus prior methods, while also supporting compositional scene reconstruction by union of independent objects.

Significance. If the empirical claims are substantiated, the work would offer a practical alternative to decoupled canonical-then-rotate pipelines in 3D generation, with direct benefits for pose-consistent assets and multi-object scenes. The absence of parameter-free derivations or machine-checked proofs is noted, but the architectural shift away from canonical assumptions is a clear conceptual contribution if the anchor mechanism proves robust.

major comments (2)

[Abstract and Experiments section] Abstract and Experiments section: the claim of 'superior geometric alignment and image-to-3D correspondence' is asserted without any reported quantitative metrics, tables, or error bars; this leaves the central performance claim unsupported and prevents assessment of whether gains exceed depth-estimation noise.
[Method description (unprojection step)] Method description (unprojection step): treating the unprojected monocular depth as a 'rigorous spatial supervision' anchor assumes metric accuracy and completeness, yet no ablation isolates the impact of known monocular depth artifacts (scale drift, boundary errors, textureless regions) on final 3D output or correspondence metrics.

minor comments (2)

Clarify the exact conditioning mechanism (e.g., how the partial point cloud is encoded and injected into the diffusion U-Net) with a diagram or pseudocode for reproducibility.
The extension to compositional scenes via simple union is promising but requires explicit discussion of overlap handling or collision resolution to be fully convincing.

Simulated Author's Rebuttal

2 responses · 0 unresolved

Thank you for the constructive feedback on our manuscript. We have carefully considered each major comment and provide our responses below, along with planned revisions to strengthen the paper.

read point-by-point responses

Referee: [Abstract and Experiments section] Abstract and Experiments section: the claim of 'superior geometric alignment and image-to-3D correspondence' is asserted without any reported quantitative metrics, tables, or error bars; this leaves the central performance claim unsupported and prevents assessment of whether gains exceed depth-estimation noise.

Authors: We agree that quantitative metrics are necessary to substantiate the claims of superior performance. The original experiments focused on qualitative results to highlight the visual improvements in pose alignment and correspondence. In the revised version, we have incorporated quantitative evaluations, including tables with metrics for geometric alignment (e.g., Chamfer distance) and image-to-3D correspondence accuracy, complete with error bars from repeated experiments. These metrics confirm that the observed gains are robust and not merely artifacts of depth estimation noise. revision: yes
Referee: [Method description (unprojection step)] Method description (unprojection step): treating the unprojected monocular depth as a 'rigorous spatial supervision' anchor assumes metric accuracy and completeness, yet no ablation isolates the impact of known monocular depth artifacts (scale drift, boundary errors, textureless regions) on final 3D output or correspondence metrics.

Authors: We acknowledge that monocular depth unprojection can introduce artifacts such as scale drift, boundary errors, and inaccuracies in textureless regions, and that our original submission did not include ablations isolating their effects. To address this, we have added a dedicated ablation study in the revised manuscript. This study systematically perturbs the depth estimates with realistic artifacts and evaluates the impact on the final 3D outputs and correspondence metrics. The results indicate that while performance degrades with severe artifacts, PAD's observation-space approach still outperforms prior methods, demonstrating the anchor's utility even under imperfect conditions. revision: yes

Circularity Check

0 steps flagged

No circularity: architectural design choice with no self-referential reduction

full rationale

The paper presents PAD as a novel end-to-end diffusion framework whose core mechanism is the explicit injection of an unprojected monocular depth point cloud as a 3D geometric anchor inside the observation space. No equations, loss terms, or derivation steps are shown that reduce the claimed geometric alignment or pose resolution to fitted parameters, self-definitions, or prior self-citations. The superiority claims rest on the architectural decision to abandon canonical-then-rotate pipelines rather than on any mathematical identity or renamed empirical pattern. The framework is therefore self-contained as a methodological proposal; external validation would require empirical comparison, not internal reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only; no explicit free parameters, axioms, or invented entities are stated. The approach implicitly relies on the accuracy of monocular depth estimation, which is treated as an external input rather than a new entity.

pith-pipeline@v0.9.0 · 5449 in / 1074 out tokens · 18627 ms · 2026-05-09T20:10:43.817877+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

67 extracted references · 29 canonical work pages · 9 internal anchors

[1]

In: 2025 International Conference on Pose-Aware Diffusion for 3D Generation 15 3D Vision (3DV)

Ardelean, A., Özer, M., Egger, B.: Gen3dsr: Generalizable 3d scene reconstruction via divide and conquer from a single view. In: 2025 International Conference on Pose-Aware Diffusion for 3D Generation 15 3D Vision (3DV). pp. 616–626. IEEE (2025)

2025
[2]

ZoeDepth: Zero-shot Transfer by Combining Relative and Metric Depth

Bhat,S.F.,Birkl,R.,Wofk,D.,Wonka,P.,Müller,M.:Zoedepth:Zero-shottransfer by combining relative and metric depth. arXiv preprint arXiv:2302.12288 (2023)

work page internal anchor Pith review arXiv 2023
[3]

arXiv preprint arXiv:2404.19525 (2024)

Chen, L., Wang, Z., Zhou, Z., Gao, T., Su, H., Zhu, J., Li, C.: Microdreamer: Efficient 3d generation insim20 seconds by score-based iterative reconstruction. arXiv preprint arXiv:2404.19525 (2024)

work page arXiv 2024
[4]

arXiv preprint arXiv:2503.13265 (2025)

Chen, L., Zhou, Z., Zhao, M., Wang, Y., Zhang, G., Huang, W., Sun, H., Wen, J.R., Li, C.: Flexworld: Progressively expanding 3d scenes for flexiable-view synthesis. arXiv preprint arXiv:2503.13265 (2025)

work page arXiv 2025
[5]

V3d: Video diffusion models are effective 3d generators

Chen, Z., Wang, Y., Wang, F., Wang, Z., Liu, H.: V3d: Video diffusion models are effective 3d generators. arXiv preprint arXiv:2403.06738 (2024)

work page arXiv 2024
[6]

Chung, J., Lee, S., Nam, H., Lee, J., Lee, K.M.: Luciddreamer: Domain-free gener- ation of 3d gaussian splatting scenes (2023),https://arxiv.org/abs/2311.13384

work page arXiv 2023
[7]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Collins, J., Goel, S., Deng, K., Luthra, A., Xu, L., Gundogdu, E., Zhang, X., Vicente, T.F.Y., Dideriksen, T., Arora, H., et al.: Abo: Dataset and benchmarks for real-world 3d object understanding. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 21126–21136 (2022)

2022
[8]

Advances in Neural Information Processing Systems36, 35799–35813 (2023)

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 universe of 10m+ 3d objects. Advances in Neural Information Processing Systems36, 35799–35813 (2023)

2023
[9]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Deitke, M., Schwenk, D., Salvador, J., Weihs, L., Michel, O., VanderBilt, E., Schmidt, L., Ehsani, K., Kembhavi, A., Farhadi, A.: Objaverse: A universe of annotated 3d objects. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 13142–13153 (2023)

2023
[10]

In: 2022 International Conference on Robotics and Automation (ICRA)

Downs, L., Francis, A., Koenig, N., Kinman, B., Hickman, R., Reymann, K., McHugh, T.B., Vanhoucke, V.: Google scanned objects: A high-quality dataset of 3d scanned household items. In: 2022 International Conference on Robotics and Automation (ICRA). pp. 2553–2560. Ieee (2022)

2022
[11]

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

Fu, H., Cai, B., Gao, L., Zhang, L.X., Wang, J., Li, C., Zeng, Q., Sun, C., Jia, R., Zhao, B., et al.: 3d-front: 3d furnished rooms with layouts and semantics. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 10933–10942 (2021)

2021
[12]

International Journal of Computer Vision129(12), 3313–3337 (2021)

Fu, H., Jia, R., Gao, L., Gong, M., Zhao, B., Maybank, S., Tao, D.: 3d-future: 3d furniture shape with texture. International Journal of Computer Vision129(12), 3313–3337 (2021)

2021
[13]

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

Han, H., Yang, R., Liao, H., Xing, J., Xu, Z., Yu, X., Zha, J., Li, X., Li, W.: Reparo: Compositional 3d assets generation with differentiable 3d layout alignment. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 25367–25377 (2025)

2025
[14]

LRM: Large Reconstruction Model for Single Image to 3D

Hong, Y., Zhang, K., Gu, J., Bi, S., Zhou, Y., Liu, D., Liu, F., Sunkavalli, K., Bui, T., Tan, H.: Lrm: Large reconstruction model for single image to 3d. arXiv preprint arXiv:2311.04400 (2023)

work page internal anchor Pith review arXiv 2023
[15]

Cupid: Pose-grounded genera- tive 3d reconstruction from a single image.arXiv preprint arXiv:2510.20776, 2025

Huang, B., Duan, H., Zhao, Y., Zhao, Z., Ma, Y., Gao, S.: Cupid: Generative 3d reconstruction via joint object and pose modeling. arXiv preprint arXiv:2510.20776 (2025)

work page arXiv 2025
[16]

In: Proceedings of the Computer Vision and Pattern Recognition Conference

Huang, Z., Guo, Y.C., An, X., Yang, Y., Li, Y., Zou, Z.X., Liang, D., Liu, X., Cao, Y.P., Sheng, L.: Midi: Multi-instance diffusion for single image to 3d scene genera- tion. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 23646–23657 (2025) 16 Zhou et al

2025
[17]

arXiv2506.15442(2025) 10

Hunyuan3D, T., Yang, S., Yang, M., Feng, Y., Huang, X., Zhang, S., He, Z., Luo, D., Liu, H., Zhao, Y., et al.: Hunyuan3d 2.1: From images to high-fidelity 3d assets with production-ready pbr material. arXiv preprint arXiv:2506.15442 (2025)

work page arXiv 2025
[18]

arXiv preprint arXiv:2602.03907 (2026)

Hunyuan3D, T., Zhang, B., Guo, C., Guo, D., Liu, H., Yan, H., Shi, H., Yu, J., Xu, J., Huang, J., et al.: Hy3d-bench: Generation of 3d assets. arXiv preprint arXiv:2602.03907 (2026)

work page arXiv 2026
[19]

ACM Transactions on Graphics42(4) (2023)

Kerbl, B., Kopanas, G., Leimkühler, T., Drettakis, G.: 3d gaussian splatting for real-time radiance field rendering. ACM Transactions on Graphics42(4) (2023)

2023
[20]

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

Khanna, M., Mao, Y., Jiang, H., Haresh, S., Shacklett, B., Batra, D., Clegg, A., Undersander, E., Chang, A.X., Savva, M.: Habitat synthetic scenes dataset (hssd- 200): An analysis of 3d scene scale and realism tradeoffs for objectgoal navigation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 16384–16393 (2024)

2024
[21]

arXiv preprint arXiv:2502.06608 (2025)

Li, Y., Zou, Z.X., Liu, Z., Wang, D., Liang, Y., Yu, Z., Liu, X., Guo, Y.C., Liang, D., Ouyang, W., et al.: Triposg: High-fidelity 3d shape synthesis using large-scale rectified flow models. arXiv preprint arXiv:2502.06608 (2025)

work page arXiv 2025
[22]

arXiv preprint arXiv:2412.12091 (2024)

Liang, H., Cao, J., Goel, V., Qian, G., Korolev, S., Terzopoulos, D., Plataniotis, K.N., Tulyakov, S., Ren, J.: Wonderland: Navigating 3d scenes from a single image. arXiv preprint arXiv:2412.12091 (2024)

work page arXiv 2024
[23]

arXiv preprint arXiv:2311.11284 (2023)

Liang, Y., Yang, X., Lin, J., Li, H., Xu, X., Chen, Y.: Luciddreamer: Towards high-fidelity text-to-3d generation via interval score matching. arXiv preprint arXiv:2311.11284 (2023)

work page arXiv 2023
[24]

Flow Matching for Generative Modeling

Lipman, Y., Chen, R.T., Ben-Hamu, H., Nickel, M., Le, M.: Flow matching for generative modeling. arXiv preprint arXiv:2210.02747 (2022)

work page internal anchor Pith review Pith/arXiv arXiv 2022
[25]

Syncdreamer: Gen- erating multiview-consistent images from a single-view im- age.arXiv preprint arXiv:2309.03453, 2023

Liu, Y., Lin, C., Zeng, Z., Long, X., Liu, L., Komura, T., Wang, W.: Syncdreamer: Generating multiview-consistent images from a single-view image. arXiv preprint arXiv:2309.03453 (2023)

work page arXiv 2023
[26]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Long, X., Guo, Y.C., Lin, C., Liu, Y., Dou, Z., Liu, L., Ma, Y., Zhang, S.H., Habermann, M., Theobalt, C., et al.: Wonder3d: Single image to 3d using cross- domain diffusion. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 9970–9980 (2024)

2024
[27]

In: Seminal graphics: pioneering efforts that shaped the field, pp

Lorensen, W.E., Cline, H.E.: Marching cubes: A high resolution 3d surface con- struction algorithm. In: Seminal graphics: pioneering efforts that shaped the field, pp. 347–353 (1998)

1998
[28]

arXiv preprint arXiv:2508.15769 (2025)

Meng, Y., Wu, H., Zhang, Y., Xie, W.: Scenegen: Single-image 3d scene generation in one feedforward pass. arXiv preprint arXiv:2508.15769 (2025)

work page arXiv 2025
[29]

Commu- nications 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. Commu- nications of the ACM65(1), 99–106 (2021)

2021
[30]

ACM Transactions on Graphics (ToG)41(4), 1– 15 (2022)

Müller,T.,Evans,A.,Schied,C.,Keller,A.:Instantneuralgraphicsprimitiveswith a multiresolution hash encoding. ACM Transactions on Graphics (ToG)41(4), 1– 15 (2022)

2022
[31]

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

Nie, Y., Han, X., Guo, S., Zheng, Y., Chang, J., Zhang, J.J.: Total3dunderstanding: Joint layout, object pose and mesh reconstruction for indoor scenes from a single image. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 55–64 (2020)

2020
[32]

Peebles,W.,Xie,S.:Scalablediffusionmodelswithtransformers.In:Proceedingsof the IEEE/CVF international conference on computer vision. pp. 4195–4205 (2023)

2023
[33]

SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis

Podell, D., English, Z., Lacey, K., Blattmann, A., Dockhorn, T., Müller, J., Penna, J., Rombach, R.: Sdxl: Improving latent diffusion models for high-resolution image synthesis. arXiv preprint arXiv:2307.01952 (2023) Pose-Aware Diffusion for 3D Generation 17

work page internal anchor Pith review arXiv 2023
[34]

DreamFusion: Text-to-3D using 2D Diffusion

Poole, B., Jain, A., Barron, J.T., Mildenhall, B.: Dreamfusion: Text-to-3d using 2d diffusion. arXiv preprint arXiv:2209.14988 (2022)

work page internal anchor Pith review arXiv 2022
[35]

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

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)

2022
[36]

Seed,B.:Seed3d1.0:Fromimagestohigh-fidelitysimulation-ready3dassets(2025)

2025
[37]

Shi, R., Chen, H., Zhang, Z., Liu, M., Xu, C., Wei, X., Chen, L., Zeng, C., Su, H.: Zero123++: a single image to consistent multi-view diffusion base model (2023)

2023
[38]

MVDream: Multi-view Diffusion for 3D Generation

Shi,Y.,Wang,P.,Ye,J.,Long,M.,Li,K.,Yang,X.:Mvdream:Multi-viewdiffusion for 3d generation. arXiv preprint arXiv:2308.16512 (2023)

work page internal anchor Pith review arXiv 2023
[39]

arXiv preprint arXiv:2601.11514 (2026)

Siddiqui, Y., Frost, D., Aroudj, S., Avetisyan, A., Howard-Jenkins, H., DeTone, D., Moulon, P., Wu, Q., Li, Z., Straub, J., et al.: Shaper: Robust conditional 3d shape generation from casual captures. arXiv preprint arXiv:2601.11514 (2026)

work page arXiv 2026
[40]

In: European Conference on Computer Vision

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. Springer (2024)

2024
[41]

arXiv preprint arXiv:2309.16653 , year=

Tang, J., Ren, J., Zhou, H., Liu, Z., Zeng, G.: Dreamgaussian: Generative gaussian splatting for efficient 3d content creation. arXiv preprint arXiv:2309.16653 (2023)

work page arXiv 2023
[42]

Team, S.D., Chen, X., Chu, F.J., Gleize, P., Liang, K.J., Sax, A., Tang, H., Wang, W., Guo, M., Hardin, T., Li, X., Lin, A., Liu, J., Ma, Z., Sagar, A., Song, B., Wang, X., Yang, J., Zhang, B., Dollár, P., Gkioxari, G., Feiszli, M., Malik, J.: Sam 3d: 3dfy anything in images (2025),https://arxiv.org/abs/2511.16624

work page internal anchor Pith review arXiv 2025
[43]

In: European Conference on Computer Vision

Voleti, V., Yao, C.H., Boss, M., Letts, A., Pankratz, D., Tochilkin, D., Laforte, C., Rombach, R., Jampani, V.: Sv3d: Novel multi-view synthesis and 3d genera- tion from a single image using latent video diffusion. In: European Conference on Computer Vision. pp. 439–457. Springer (2024)

2024
[44]

arXiv preprint arXiv:2511.23191 (2025)

Wan, Y., Liu, L., Zhou, J., Zhou, Z., Zhang, X., Zhang, D., Jiao, S., Hou, Q., Cheng, M.M.: Geoworld: Unlocking the potential of geometry models to facilitate high-fidelity 3d scene generation. arXiv preprint arXiv:2511.23191 (2025)

work page arXiv 2025
[45]

In: Proceedings of the Computer Vision and Pattern Recognition Conference

Wang, J., Chen, M., Karaev, N., Vedaldi, A., Rupprecht, C., Novotny, D.: Vggt: Visual geometry grounded transformer. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 5294–5306 (2025)

2025
[46]

arXiv preprint arXiv:2312.02201 , year=

Wang, P., Shi, Y.: Imagedream: Image-prompt multi-view diffusion for 3d genera- tion. arXiv preprint arXiv:2312.02201 (2023)

work page arXiv 2023
[47]

In: Proceedings of the Computer Vision and Pattern Recog- nition Conference

Wang, R., Xu, S., Dai, C., Xiang, J., Deng, Y., Tong, X., Yang, J.: Moge: Unlock- ing accurate monocular geometry estimation for open-domain images with optimal training supervision. In: Proceedings of the Computer Vision and Pattern Recog- nition Conference. pp. 5261–5271 (2025)

2025
[48]

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

Wang, S., Leroy, V., Cabon, Y., Chidlovskii, B., Revaud, J.: Dust3r: Geometric 3d vision made easy. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 20697–20709 (2024)

2024
[49]

Advances in Neural Information Processing Systems36(2024)

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

2024
[50]

In: ACM SIGGRAPH 2024 Conference Papers

Wang, Z., Yuan, Z., Wang, X., Li, Y., Chen, T., Xia, M., Luo, P., Shan, Y.: Mo- tionctrl: A unified and flexible motion controller for video generation. In: ACM SIGGRAPH 2024 Conference Papers. pp. 1–11 (2024)

2024
[51]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Wen, B., Yang, W., Kautz, J., Birchfield, S.: Foundationpose: Unified 6d pose esti- mation and tracking of novel objects. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 17868–17879 (2024) 18 Zhou et al

2024
[52]

Tech report (2025)

Xiang, J., Chen, X., Xu, S., Wang, R., Lv, Z., Deng, Y., Zhu, H., Dong, Y., Zhao, H., Yuan, N.J., Yang, J.: Native and compact structured latents for 3d generation. Tech report (2025)

2025
[53]

In: Proceedings of the Computer Vision and Pattern Recognition Conference

Xiang, J., Lv, Z., Xu, S., Deng, Y., Wang, R., Zhang, B., Chen, D., Tong, X., Yang, J.: Structured 3d latents for scalable and versatile 3d generation. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 21469–21480 (2025)

2025
[54]

Camco: Camera-controllable 3d-consistent image-to-video generation.arXiv preprint arXiv:2406.02509, 2024

Xu, D., Nie, W., Liu, C., Liu, S., Kautz, J., Wang, Z., Vahdat, A.: Camco: Camera-controllable 3d-consistent image-to-video generation. arXiv preprint arXiv:2406.02509 (2024)

work page arXiv 2024
[55]

InstantMesh: Efficient 3D Mesh Generation from a Single Image with Sparse-view Large Reconstruction Models

Xu, J., Cheng, W., Gao, Y., Wang, X., Gao, S., Shan, Y.: Instantmesh: Efficient 3d mesh generation from a single image with sparse-view large reconstruction models. arXiv preprint arXiv:2404.07191 (2024)

work page internal anchor Pith review arXiv 2024
[56]

In: European Conference on Computer Vision

Xu, Y., Shi, Z., Yifan, W., Chen, H., Yang, C., Peng, S., Shen, Y., Wetzstein, G.: Grm: Large gaussian reconstruction model for efficient 3d reconstruction and gen- eration. In: European Conference on Computer Vision. pp. 1–20. Springer (2024)

2024
[57]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Xue, L., Gao, M., Xing, C., Martín-Martín, R., Wu, J., Xiong, C., Xu, R., Niebles, J.C., Savarese, S.: Ulip: Learning a unified representation of language, images, and point clouds for 3d understanding. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 1179–1189 (2023)

2023
[58]

ACM Transactions on Graphics (TOG)44(4), 1–19 (2025)

Yao, K., Zhang, L., Yan, X., Zeng, Y., Zhang, Q., Xu, L., Yang, W., Gu, J., Yu, J.: Cast: Component-aligned 3d scene reconstruction from an rgb image. ACM Transactions on Graphics (TOG)44(4), 1–19 (2025)

2025
[59]

Freeman, and Jiajun Wu

Yu, H.X., Duan, H., Herrmann, C., Freeman, W.T., Wu, J.: Wonderworld: Inter- active 3d scene generation from a single image. arXiv preprint arXiv:2406.09394 (2024)

work page arXiv 2024
[60]

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

Yu, H.X., Duan, H., Hur, J., Sargent, K., Rubinstein, M., Freeman, W.T., Cole, F., Sun, D., Snavely, N., Wu, J., et al.: Wonderjourney: Going from anywhere to everywhere. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 6658–6667 (2024)

2024
[61]

Yu, W., Xing, J., Yuan, L., Hu, W., Li, X., Huang, Z., Gao, X., Wong, T.T., Shan, Y., Tian, Y.: Viewcrafter: Taming video diffusion models for high-fidelity novel view synthesis (2024),https://arxiv.org/abs/2409.02048

work page internal anchor Pith review arXiv 2024
[62]

arXiv preprint arXiv:2501.05763 (2025)

Zhai, S., Ye, Z., Liu, J., Xie, W., Hu, J., Peng, Z., Xue, H., Chen, D., Wang, X., Yang, L., et al.: Stargen: A spatiotemporal autoregression framework with video diffusion model for scalable and controllable scene generation. arXiv preprint arXiv:2501.05763 (2025)

work page arXiv 2025
[63]

ACM Transactions On Graphics (TOG)42(4), 1–16 (2023)

Zhang, B., Tang, J., Niessner, M., Wonka, P.: 3dshape2vecset: A 3d shape repre- sentation for neural fields and generative diffusion models. ACM Transactions On Graphics (TOG)42(4), 1–16 (2023)

2023
[64]

ACM Transactions on Graphics (TOG)43(4), 1–20 (2024)

Zhang, L., Wang, Z., Zhang, Q., Qiu, Q., Pang, A., Jiang, H., Yang, W., Xu, L., Yu, J.: Clay: A controllable large-scale generative model for creating high-quality 3d assets. ACM Transactions on Graphics (TOG)43(4), 1–20 (2024)

2024
[65]

Zhao, Q., Zhang, X., Xu, H., Chen, Z., Xie, J., Gao, Y., Tu, Z.: Depr: Depth guided single-viewscenereconstructionwithinstance-leveldiffusion.In:Proceedingsofthe IEEE/CVF International Conference on Computer Vision. pp. 5722–5733 (2025)

2025
[66]

Advances in Neural Information Processing Systems37, 39104–39127 (2024)

Zhou, J., Liu, Y.S., Han, Z.: Zero-shot scene reconstruction from single images with deep prior assembly. Advances in Neural Information Processing Systems37, 39104–39127 (2024)

2024
[67]

Zhou, J., Wang, J., Ma, B., Liu, Y.S., Huang, T., Wang, X.: Uni3d: Exploring unified 3d representation at scale. arXiv preprint arXiv:2310.06773 (2023) Pose-Aware Diffusion for 3D Generation 19 A More implementation details A.1 Training data curation We construct our training set primarily from the Objaverse [9] and 3D-Front [11] datasets.Toensurehighgeom...

work page arXiv 2023