Pith. sign in

REVIEW 5 major objections 5 minor 39 references

MVBoost: Boost 3D Reconstruction with Multi-View Refinement

T0 review · 5 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read MVBoost constructs pseudo-ground-truth multi-view images by refining renders of a 3D model through a diffusion model, and training on that synthetic data yields higher-fidelity single-image 3D reconstruction than prior open-world methods…

desk verdict A plausible self-training recipe for 3D reconstruction whose headline numbers rest on unvalidated pseudo-GT and a hyperparameter tuned on the test benchmark. read the letter →

arxiv 2411.17772 v2 pith:KGUGJK3N submitted 2024-11-26 cs.CV cs.AI

Xiangyu Liu , Xiaomei Zhang , Zhiyuan Ma , Xiangyu Zhu , Zhen Lei This is my paper · ORCID
classification cs.CVcs.AI
keywords single-image3Dreconstructionmulti-viewdiffusionpseudo-ground-truthgenerationGaussianSplattingLoRAfine-tuningsynthetictrainingdatanovelviewsynthesisGoogleScannedObjects
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

MVBoost aims to solve a bottleneck in single-image 3D reconstruction: the scarcity of diverse, high-quality 3D training data. It constructs pseudo-ground-truth multi-view images by first generating multiple views of an input image with a multi-view diffusion model, then passing those views through a large 3D reconstruction model to get a consistent 3D Gaussian Splatting, and finally partially denoising renderings of that splatting to produce refined views. These refined views supervise a feed-forward reconstruction model, augmented with LoRA, so the model learns from a large synthetic dataset without any new 3D scans. On the Google Scanned Objects benchmark the paper reports gains across rendering quality and geometry metrics, including PSNR 18.561, SSIM 0.859, LPIPS 0.131 and Chamfer Distance 0.10110, and it shows the same recipe also improves OpenLRM.

What carries the argument

The central mechanism is the multi-view refinement strategy: a forward diffusion step adds Gaussian noise with strength $s = 0.95$ to renderings of the 3D Gaussian Splatting produced by the base reconstruction model, and a reverse diffusion step conditioned on the original input view partially denoises them into refined multi-view images. These refined images serve as pseudo-ground truth to supervise the boosted reconstruction model through a combined MSE and LPIPS loss, while LoRA (low-rank adaptation) inserted into the cross-view self-attention keeps training stable. A separate input view optimization post-process searches over camera poses with LPIPS loss and then optimizes a learnable additive matrix on the splatting, so the most important viewpoint matches the user's input.

What would settle it

Measure the multi-view consistency of the refined pseudo-ground-truth set directly, for example by computing pair-wise depth reprojection error between neighboring refined views. If the refined set is less consistent than the raw multi-view output, or if the reported PSNR and Chamfer Distance improvements disappear when the base reconstruction model is swapped for a different backbone, the central claim would be undercut.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes that a closed loop between a multi-view diffusion model and a 3D reconstruction model can generate useful training data for feed-forward 3D reconstruction. The loop works because the two components fail differently: the diffusion model produces high-fidelity but cross-view inconsistent images, while the reconstruction model produces geometrically consistent but blurrier renders. By rendering the consistent 3D model and partially denoising those renders at a carefully chosen noise strength, the method produces a refined multi-view set that is both detailed and coherent. Training a reconstruction model with LoRA on this synthetic dataset, rather than on existing 3D assets, yields higher-quality geometry and renderings than prior open-world methods on the GSO benchmark.

Load-bearing premise

The load-bearing assumption is that the partially denoised renderings of the base 3D model are both more accurate than the raw multi-view outputs and consistent enough across views to act as reliable ground truth; the paper tunes the noise strength but never directly measures the cross-view consistency of the refined set.

Editorial extensions

If this is right

  • A feed-forward single-image-to-3D model can be trained without any 3D ground truth, using only single-view images plus diffusion refinement, and still outperform models trained on existing 3D datasets.
  • The same pseudo-ground-truth loop transfers to other reconstruction architectures, since boosting OpenLRM improves its PSNR from 16.728 to 17.023 and its F-Score from 0.6562 to 0.6832.
  • The refinement strength acts as a tunable hyperparameter: noise 0.95 gives the best training signal, with PSNR rising from 17.811 (original) to 19.132 (refined) in the 2D ablation.
  • Input view optimization can drive the LPIPS of the aligned view from 0.108 to 0.002, so the final asset can be tailored to the exact view a user provides.
  • Because the dataset is synthesized from text-to-image prompts, acquiring training images is nearly cost-free and can be expanded to new object categories on demand.

Reading between the lines

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

  • If the refinement loop is trustworthy, it suggests that 3D reconstruction models can be continuously improved on new image distributions without recapturing 3D data; one test would be to run the loop on user-uploaded photos and measure whether reconstruction quality tracks image-domain diversity.
  • The paper does not directly measure cross-view consistency of the refined multi-view set; a natural extension would be to report pair-wise reprojection error before and after refinement, since the claim that consistency is preserved is load-bearing.
  • The input view optimization is a per-asset post-process; combining it with a lightweight pose predictor at inference time could remove the search over camera poses for real-time use.
  • Because LoRA is applied only in attention layers, the boost may partly come from adapting multi-view fusion behavior rather than from new geometric reasoning; ablating LoRA placement would clarify the mechanism.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 5 minor

Summary. The paper proposes MVBoost, a framework that constructs pseudo-ground-truth multi-view data by first generating multi-views from a single image with a multi-view diffusion model (Era3D), reconstructing a 3D Gaussian Splatting with a large reconstruction model (LGM), rendering these views, and then partially denoising noisy renders with the same diffusion model. The resulting refined multi-view set is used to supervise a LoRA-tuned version of the base reconstruction model via MSE plus LPIPS loss. The paper also introduces an input-view optimization post-processing step that aligns the reconstructed 3D asset with the user-provided view. Experiments on the Google Scanned Objects (GSO) benchmark report state-of-the-art PSNR 18.561, SSIM 0.859, LPIPS 0.131 (Table 1) and Chamfer distance 0.10110 with F-Score 0.7977 (Table 2).

Significance. If the central claim holds, MVBoost offers a general recipe for boosting feed-forward single-image-to-3D models using synthetic multi-view data without collecting new 3D ground truth. The method is shown to be architecture-agnostic in a limited sense, as an experiment with OpenLRM (Tables 5-6) reproduces the boosting effect on a different reconstruction backbone. The training cost is low (about one day on 8 A100s), and a code repository is provided. However, the paper's main evidence is fragile: the pseudo-ground-truth quality is not directly measured, the key refinement-strength hyperparameter is selected on a subset of the evaluation benchmark, and the headline results are presented without error bars. These issues prevent the significance from being fully established.

major comments (5)
  1. [§3.2, Eq. (5); §4.3, Tables 3-4] The pseudo-ground-truth set C_pi_up is the sole new supervision signal in Eq. (6), yet the paper never directly validates that C_pi_up is more accurate or more cross-view consistent than the raw multi-view output C_pi. Table 3, if it is intended to measure refinement quality against ground-truth renders, does not say so, and the caption describes it only as an 'ablation of the multi-view refinement strategy' without specifying the reference. Even if Table 3 is meant as this validation, it is computed on a 300-asset subset of GSO that is also used to select s, so it cannot act as independent evidence. The authors should report, on a held-out split, PSNR/LPIPS of C_pi_up versus ground-truth renders and a cross-view consistency metric (e.g., pairwise LPIPS or geometric reprojection error), alongside the raw C_pi values.
  2. [§4.3, Tables 3-4 vs. §4.2, Tables 1-2] The refinement strength s=0.95 is selected by an ablation on a randomly chosen 300-asset subset of GSO (Table 3) and then used to produce the headline results on the full GSO benchmark (Tables 1-2). Since the 300-asset subset is a subset of the test set, the main comparisons are not independent of hyperparameter selection. The authors should either use a proper train/validation split disjoint from evaluation, or report results across several values of s and across random seeds with error bars, so the reader can assess the stability of the claimed margin.
  3. [§3.2, Eqs. (4)-(5)] The mechanistic claim that the refinement 'combines the high accuracy of the multi-view generation model and the consistency of the 3D reconstruction model' is not supported at the chosen noise strength. At s=0.95, t=0.95T, the forward-process signal contribution from the render x_pi is heavily attenuated, so the backward diffusion is dominated by the multi-view diffusion prior conditioned on the input image c. The paper provides no analysis showing that the 3D-consistent render meaningfully influences the output at this strength. The authors should quantify the information flow from x_pi into C_pi_up, e.g., by ablating s at lower values and measuring how much the refinement output changes when x_pi is replaced by random noise, or by reporting the signal-to-noise ratio at the selected t.
  4. [§3.3, §4.1] The paper repeatedly states that the method 'eliminates the need for 3D datasets' and that 'we do not rely on any existing image datasets for training.' This is misleading: the base reconstruction model R_phi (LGM) and the multi-view diffusion model (Era3D) are pretrained on large 3D and multi-view datasets, and the proposed pipeline only adds a LoRA fine-tuning stage. What the method avoids is collecting new 3D ground truth for the boosting stage, which is a legitimate but narrower contribution. The text should be revised to state this scope precisely, and the abstract's wording 'without collecting new 3D ground truth' is the accurate formulation.
  5. [§4.2, Tables 1-2] The main quantitative comparisons contain no error bars, confidence intervals, or significance tests. Since the GSO benchmark contains a large number of objects, the authors should either report standard deviations over object subsets or run repeated evaluations with different random seeds. Without such statistics, the reported margins over the second-best method (e.g., PSNR 18.561 vs. 17.435, CD 0.10110 vs. 0.12132) cannot be assessed for robustness.
minor comments (5)
  1. [Throughout] There are several typos and formatting issues: 'VFsuion3D' in Tables 1, 2, 5 and 6; 'Disscussion' in the Section 4.3 heading; a stray 'D' before 'The input view optimization' in the ablation section; 'The rendering process can be be represented' in Section 3.1; and inconsistent citation formatting (e.g., [29] is listed both as a published ECCV paper and as a preprint in the reference list). A careful proofread is needed.
  2. [§4.3, Tables 3 and 4] The captions of Tables 3 and 4 do not state what is being measured or what the reference is (e.g., ground-truth renders from GSO, or rendered views of the reconstructed 3D model). The reader must infer the evaluation protocol, which is especially problematic because Table 3 is used to justify the central hyperparameter choice. The captions should be expanded to specify the metric reference and the dataset split.
  3. [§3.4, Eq. (7)] The input-view optimization searches over the 'set of all possible camera poses' but no details are given about the pose search space, initialization, or the number of optimization steps. The reported LPIPS drop from 0.108 to 0.002 on the input view is expected because the method optimizes directly against that view; the paper should also report the effect on other views to demonstrate that this post-processing does not degrade global consistency.
  4. [§4.1, Datasets] The description of the training-data generation is vague: the authors say they 'employ ChatGPT to generate over 100k prompts' and use a text-to-image model, but do not specify which text-to-image model is used, whether the generated images are filtered, or how the single-view dataset relates to the evaluation domain. This information is necessary for reproducibility.
  5. [§7, Algorithm 1] In Algorithm 1, line 6 sets t=sT, but the values of s are given only in the ablation tables; the algorithm should explicitly state the range of s and the mapping from s to the noise schedule (alpha_t, sigma_t) used in Eq. (4).

Circularity Check

1 steps flagged · score 2.0 of 10

No central circularity: the GSO improvement is an externally evaluated distillation result; one minor self-referential evaluation in Input View Optimization.

  1. other [Section 3.4 (Eqs. 7 and 9) and Section 4.3 (reported LPIPS decrease).]
    "The matrix W is optimized by minimizing the LPIPS loss between the rendered view at the optimal camera pose πopt and the input view c: W = arg min_W LPIPS(g(θ↑, πopt), c). ... The input view optimization is a post-processing method. All previous quantitative and qualitative experiments are conducted without Input View Optimization. ... On the GSO dataset, the LPIPS loss of the input view decreased from an average of 0.108 to 0.002."

    The reported evidence for the Input View Optimization module is the LPIPS decrease from 0.108 to 0.002. But πopt and W are defined as exact minimizers of LPIPS in Eqs. (7) and (9), so this decrease is a mathematical consequence of solving the same objective, not an independent measurement. The paper is transparent that this is post-processing and excludes it from Tables 1 and 2, so the central SOTA claim does not rest on this step; the circularity is confined to this self-evaluation.

full rationale

The central claim of MVBoost is that training LoRA-augmented R_phi* to match the refined multi-view set C_pi_up improves single-image reconstruction relative to external GSO ground truth. This is not circular by construction: C_pi_up is produced by a fixed teacher (multi-view diffusion G plus base reconstruction R_phi) through Eqs. (4)-(5), while the student R_phi* is a different LoRA parameterization optimized in Eq. (6). The targets are not the student's own outputs, and the reported PSNR/SSIM/LPIPS/CD/F-Score are computed against independent GSO meshes and renders, not against C_pi_up. The pseudo-GT loop is a form of self-distillation, which can be a legitimate training scheme; whether the unvalidated C_pi_up is actually reliable (especially at s=0.95, where the noise is near-pure and the 'consistency' signal from x_pi is weak) is an evidentiary/correctness concern, not an equation-level equivalence. The noise-strength sweep on a 300-asset subset is hyperparameter selection, not a fitted parameter renamed as a prediction. There are no load-bearing self-citations or uniqueness-theorem imports. The only concrete circularity is the minor self-referential evaluation of Input View Optimization, which is disclosed as post-processing and excluded from the main comparisons; hence score 2 rather than 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical or model-level entities. The trainable matrix W in Eq. (8) is a per-instance optimization variable, not an invented entity. The main dependency is on the base models and their pretraining, plus the tuned refinement strength.

free parameters (3)
  • Refine strength s = 0.95
    Chosen from ablation on a 300-asset GSO subset (Table 3), then used for the full GSO evaluation (Table 1). This is a test-set-tuned hyperparameter controlling the noise scale in the diffusion refinement (Eqs. 4-5).
  • LoRA rank = 32
    Stated in the supplementary with no ablation. Controls the capacity of the boosted model.
  • Supervision viewpoint set = six fixed poses (front, front right, right, back, left, front left)
    The fixed set used for rendering supervision could be tuned; no ablation on the number or choice of views is provided.
assumptions (4)
  • domain assumption Pre-trained multi-view diffusion (Era3D) and reconstruction model (LGM) provide complementary strengths; rendering from LGM gives cross-view consistency, and Era3D provides accuracy, so their combination yields reliable pseudo-GT.
    Section 3.2. This is the core premise for the data generation, but no direct measurement of the consistency vs accuracy trade-off is provided beyond final metrics.
  • domain assumption Partial denoising with strength s improves the rendered views without breaking cross-view consistency.
    Eqs. (4)-(5) and the ablation Table 3. The paper reports PSNR/SSIM/LPIPS of the refined images but no multi-view consistency metric.
  • domain assumption Synthetic training images from ChatGPT prompts and an unspecified text-to-image model transfer to real-world GSO thumbnails.
    Section 4.1. The text-to-image model is not named and no domain-gap analysis is provided.
  • standard math Standard diffusion model forward/reverse processes and differentiable 3D Gaussian Splatting rendering are valid.
    Section 3.1 equations (1)-(3) rely on standard diffusion and 3DGS rendering operations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MVBoost: Boost 3D Reconstruction with Multi-View Refinement." pith.science (2026). https://pith.science/paper/KGUGJK3N

@misc{pith2026241117772,
  author       = {Pith},
  title        = {Pith review of: MVBoost: Boost 3D Reconstruction with Multi-View Refinement},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KGUGJK3N}},
  note         = {Machine review of arXiv:2411.17772}
}
read the original abstract

Recent advancements in 3D object reconstruction have been remarkable, yet most current 3D models rely heavily on existing 3D datasets. The scarcity of diverse 3D datasets results in limited generalization capabilities of 3D reconstruction models. In this paper, we propose a novel framework for boosting 3D reconstruction with multi-view refinement (MVBoost) by generating pseudo-GT data. The key of MVBoost is combining the advantages of the high accuracy of the multi-view generation model and the consistency of the 3D reconstruction model to create a reliable data source. Specifically, given a single-view input image, we employ a multi-view diffusion model to generate multiple views, followed by a large 3D reconstruction model to produce consistent 3D data. MVBoost then adaptively refines these multi-view images, rendered from the consistent 3D data, to build a large-scale multi-view dataset for training a feed-forward 3D reconstruction model. Additionally, the input view optimization is designed to optimize the corresponding viewpoints based on the user's input image, ensuring that the most important viewpoint is accurately tailored to the user's needs. Extensive evaluations demonstrate that our method achieves superior reconstruction results and robust generalization compared to prior works.

Figures

Figures reproduced from arXiv: 2411.17772 by the authors.

Figure 1
Figure 1. Given a single image as input, our MVBoost can generate a high-quality 3D asset. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overview of our MVBoost framework. Given a single-view image dataset, we first employ a multi-view diffusion model [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Qualitative comparison of image-to-3D methods. Our approach demonstrates superior 3D generation across a range of challeng [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Qualitative comparison of 2D multi-view data before and after refinement. Our refined multi-view shows enhanced geometric [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Our multi-view refinement strategy effectively corrects substantial viewpoint errors in the original views. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Post-processing with Input View Optimization aligns the [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Qualitative experiments of boosted OpenLRM and original OpenLRM. [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 18 canonical work pages

  1. [1]

    Multi-garment net: Learning to dress 3d people from images

    Bharat Lal Bhatnagar, Garvita Tiwari, Christian Theobalt, and Gerard Pons-Moll. Multi-garment net: Learning to dress 3d people from images. In Proceedings of the IEEE/CVF international conference on computer vision , pages 5420– 5430, 2019. 2

    2019

  2. [2]

    Chang, Thomas Funkhouser, Leonidas Guibas, Pat Hanrahan, Qixing Huang, Zimo Li, Silvio Savarese, Mano- lis Savva, Shuran Song, Hao Su, Jianxiong Xiao, Li Yi, and Fisher Yu

    Angel X. Chang, Thomas Funkhouser, Leonidas Guibas, Pat Hanrahan, Qixing Huang, Zimo Li, Silvio Savarese, Mano- lis Savva, Shuran Song, Hao Su, Jianxiong Xiao, Li Yi, and Fisher Yu. ShapeNet: An Information-Rich 3D Model Repository. Technical Report arXiv:1512.03012 [cs.GR], Stanford University — Princeton University — Toyota Tech- nological Institute at ...

    arXiv 2015

  3. [3]

    Objaverse: A universe of annotated 3d objects

    Matt Deitke, Dustin Schwenk, Jordi Salvador, Luca Weihs, Oscar Michel, Eli VanderBilt, Ludwig Schmidt, Kiana Ehsani, Aniruddha Kembhavi, and Ali Farhadi. Objaverse: A universe of annotated 3d objects. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13142–13153, 2023. 4

    2023

  4. [4]

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

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

    2024

  5. [5]

    Diffusion models beat gans on image synthesis

    Prafulla Dhariwal and Alexander Nichol. Diffusion models beat gans on image synthesis. Advances in neural informa- tion processing systems, 34:8780–8794, 2021. 4

    2021

  6. [6]

    McHugh, and Vincent Vanhoucke

    Laura Downs, Anthony Francis, Nate Koenig, Brandon Kin- man, Ryan Hickman, Krista Reymann, Thomas B. McHugh, and Vincent Vanhoucke. Google scanned objects: A high- quality dataset of 3d scanned household items, 2022. 2, 5

    work page 2022

  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]

    Gancraft: Unsupervised 3d neural rendering of minecraft worlds

    Zekun Hao, Arun Mallya, Serge Belongie, and Ming-Yu Liu. Gancraft: Unsupervised 3d neural rendering of minecraft worlds. In Proceedings of the IEEE/CVF International Con- ference on Computer Vision, pages 14072–14082, 2021. 2

    work page 2021

Show all 39 references
  1. [9]

    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. 4

    2020

  2. [10]

    LRM: Large reconstruction model for single image to 3d

    Yicong Hong, Kai Zhang, Jiuxiang Gu, Sai Bi, Yang Zhou, Difan Liu, Feng Liu, Kalyan Sunkavalli, Trung Bui, and Hao Tan. LRM: Large reconstruction model for single image to 3d. In The Twelfth International Conference on Learning Representations, 2024. 2, 3, 4, 5, 7

    2024

  3. [11]

    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. InIn- ternational Conference on Learning Representations , 2022. 2

    2022

  4. [12]

    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 Transactions on Graphics, 42 (4), 2023. 4

    2023

  5. [13]

    Instant3d: Fast text-to-3d with sparse-view generation and large reconstruction model

    Jiahao Li, Hao Tan, Kai Zhang, Zexiang Xu, Fujun Luan, Yinghao Xu, Yicong Hong, Kalyan Sunkavalli, Greg Shakhnarovich, and Sai Bi. Instant3d: Fast text-to-3d with sparse-view generation and large reconstruction model. In The Twelfth International Conference on Learning Represe...

    2024

  6. [14]

    Era3d: High-resolution multiview diffusion using efficient row-wise attention

    Peng Li, Yuan Liu, Xiaoxiao Long, Feihu Zhang, Cheng Lin, Mengfei Li, Xingqun Qi, Shanghang Zhang, Wenhan Luo, Ping Tan, et al. Era3d: High-resolution multiview diffusion using efficient row-wise attention. arXiv preprint arXiv:2405.11616, 2024. 3, 5

    2024 arXiv

  7. [15]

    High-fidelity clothed avatar reconstruction from a single image

    Tingting Liao, Xiaomei Zhang, Yuliang Xiu, Hongwei Yi, Xudong Liu, Guo-Jun Qi, Yong Zhang, Xuan Wang, Xi- angyu Zhu, and Zhen Lei. High-fidelity clothed avatar reconstruction from a single image. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognit...

    2023

  8. [16]

    One-2-3-45++: Fast single image to 3d objects with consistent multi-view generation and 3d diffusion

    Minghua Liu, Ruoxi Shi, Linghao Chen, Zhuoyang Zhang, Chao Xu, Xinyue Wei, Hansheng Chen, Chong Zeng, Ji- ayuan Gu, and Hao Su. One-2-3-45++: Fast single image to 3d objects with consistent multi-view generation and 3d diffusion. arXiv preprint arXiv:2311.07885, 2023. 3

    2023 arXiv

  9. [17]

    One-2-3-45: Any single image to 3d mesh in 45 seconds without per-shape optimiza- tion

    Minghua Liu, Chao Xu, Haian Jin, Linghao Chen, Mukund Varma T, Zexiang Xu, and Hao Su. One-2-3-45: Any single image to 3d mesh in 45 seconds without per-shape optimiza- tion. Advances in Neural Information Processing Systems , 36, 2024. 2

    2024

  10. [18]

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

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

    2023

  11. [19]

    Syncdreamer: Gen- erating multiview-consistent images from a single-view im- age

    Yuan Liu, Cheng Lin, Zijiao Zeng, Xiaoxiao Long, Lingjie Liu, Taku Komura, and Wenping Wang. Syncdreamer: Gen- erating multiview-consistent images from a single-view im- age. In The Twelfth International Conference on Learning Representations, 2024. 2, 3

    2024

  12. [20]

    Wonder3d: Single image to 3d using cross-domain diffusion

    Xiaoxiao Long, Yuan-Chen Guo, Cheng Lin, Yuan Liu, Zhiyang Dou, Lingjie Liu, Yuexin Ma, Song-Hai Zhang, Marc Habermann, Christian Theobalt, and Wenping Wang. Wonder3d: Single image to 3d using cross-domain diffusion. In 2024 IEEE/CVF Conference on Computer Vision and Pat- tern...

    2024

  13. [21]

    Scaledreamer: Scalable text-to- 3d synthesis with asynchronous score distillation

    Zhiyuan Ma, Yuxiang Wei, Yabin Zhang, Xiangyu Zhu, Zhen Lei, and Lei Zhang. Scaledreamer: Scalable text-to- 3d synthesis with asynchronous score distillation. In Euro- 9 pean Conference on Computer Vision, pages 1–19. Springer,

  14. [22]

    Self-supervised 3d shape and viewpoint estimation from single images for robotics

    Oier Mees, Maxim Tatarchenko, Thomas Brox, and Wolfram Burgard. Self-supervised 3d shape and viewpoint estimation from single images for robotics. In 2019 IEEE/RSJ Interna- tional Conference on Intelligent Robots and Systems (IROS), pages 6083–6089. IEEE, 2019. 2

    2019

  15. [23]

    Srinivasan, Matthew Tancik, Jonathan T

    Ben Mildenhall, Pratul P. Srinivasan, Matthew Tancik, Jonathan T. Barron, Ravi Ramamoorthi, and Ren Ng. Nerf: representing scenes as neural radiance fields for view synthe- sis. Commun. ACM, 65(1):99–106, 2021. 4

    2021

  16. [24]

    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. 3

    2021

  17. [25]

    Barron, and Ben Milden- hall

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

  18. [26]

    Flexible isosurface extraction for gradient-based mesh optimization

    Tianchang Shen, Jacob Munkberg, Jon Hasselgren, Kangxue Yin, Zian Wang, Wenzheng Chen, Zan Gojcic, Sanja Fidler, Nicholas Sharp, and Jun Gao. Flexible isosurface extraction for gradient-based mesh optimization. ACM Trans. Graph., 42(4), 2023. 3

    2023

  19. [27]

    Zero123++: a single image to consistent multi-view dif- fusion base model, 2023

    Ruoxi Shi, Hansheng Chen, Zhuoyang Zhang, Minghua Liu, Chao Xu, Xinyue Wei, Linghao Chen, Chong Zeng, and Hao Su. Zero123++: a single image to consistent multi-view dif- fusion base model, 2023. 3

    2023

  20. [28]

    Deep unsupervised learning using nonequilibrium thermodynamics

    Jascha Sohl-Dickstein, Eric Weiss, Niru Maheswaranathan, and Surya Ganguli. Deep unsupervised learning using nonequilibrium thermodynamics. In International confer- ence on machine learning, pages 2256–2265. PMLR, 2015. 4

    2015

  21. [29]

    Lgm: Large multi-view gaussian model for high-resolution 3d content creation

    Jiaxiang Tang, Zhaoxi Chen, Xiaokang Chen, Tengfei Wang, Gang Zeng, and Ziwei Liu. Lgm: Large multi-view gaussian model for high-resolution 3d content creation. In European Conference on Computer Vision, pages 1–18. Springer, 2025. 2, 4, 5, 7

    2025

  22. [30]

    Yeh, and Greg Shakhnarovich

    Haochen Wang, Xiaodan Du, Jiahao Li, Raymond A. Yeh, and Greg Shakhnarovich. Score jacobian chaining: Lift- ing pretrained 2d diffusion models for 3d generation. arXiv preprint arXiv:2212.00774, 2022. 3

    2022 arXiv

  23. [31]

    Imagedream: Image-prompt multi-view diffusion for 3d generation, 2023

    Peng Wang and Yichun Shi. Imagedream: Image-prompt multi-view diffusion for 3d generation, 2023. 2, 3

    2023

  24. [32]

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

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

    2024

  25. [33]

    Crm: Single image to 3d textured mesh with convolutional reconstruction model

    Zhengyi Wang, Yikai Wang, Yifei Chen, Chendong Xi- ang, Shuo Chen, Dajiang Yu, Chongxuan Li, Hang Su, and Jun Zhu. Crm: Single image to 3d textured mesh with convolutional reconstruction model. arXiv preprint arXiv:2403.05034, 2024. 2, 3, 5, 7

    2024 arXiv

  26. [34]

    Instantmesh: Efficient 3d mesh generation from a single image with sparse-view large reconstruction models

    Jiale Xu, Weihao Cheng, Yiming Gao, Xintao Wang, Shenghua Gao, and Ying Shan. Instantmesh: Efficient 3d mesh generation from a single image with sparse-view large reconstruction models. arXiv preprint arXiv:2404.07191 ,

    arXiv

  27. [35]

    Grm: Large gaus- sian reconstruction model for efficient 3d reconstruction and generation

    Yinghao Xu, Zifan Shi, Wang Yifan, Sida Peng, Ceyuan Yang, Yujun Shen, and Wetzstein Gordon. Grm: Large gaus- sian reconstruction model for efficient 3d reconstruction and generation. arxiv: 2403.14621, 2024. 4

    2024 arXiv

  28. [36]

    Triplane meets gaussian splatting: Fast and generalizable single- view 3d reconstruction with transformers

    Zi-Xin Zou, Zhipeng Yu, Yuan-Chen Guo, Yangguang Li, Ding Liang, Yan-Pei Cao, and Song-Hai Zhang. Triplane meets gaussian splatting: Fast and generalizable single- view 3d reconstruction with transformers. arXiv preprint arXiv:2312.09147, 2023. 5, 7 10 MVBoost: Boost 3D Recons...

    2023 arXiv

  29. [37]

    In our training process, we use a fixed set of six viewpoints (front, front right, right, back, left, front left) for supervision

    More Implementation Details Training. In our training process, we use a fixed set of six viewpoints (front, front right, right, back, left, front left) for supervision. Our camera model employed orthographic projection. The rank of the LoRA layer in our boosted model is 32. Me...

  30. [38]

    The al- gorithm details are presented in Algorithm 1

    More Details about Method MVBoost generates refined multi-view as pseudo-ground truth through the Multi-View Refinement Strategy. The al- gorithm details are presented in Algorithm 1. The symbols used in the algorithm are explained and defined in the main paper. Algorithm 1: M...

  31. [39]

    To further illustrate the versatility of our approach, we boost OpenLRM with multi-view refinement

    Experiment on OpenLRM The framework is compatible with various reconstruction models, supporting different types of 3D representations. To further illustrate the versatility of our approach, we boost OpenLRM with multi-view refinement. We employ a dataset of 5k refined multi-v...

Pith tools

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