MTFusion: Reconstructing Any 3D Object from Single Image Using Multi-word Textual Inversion

Jiaqi Li; Qijun Zhao; Ruowei Wang; Yu Liu; Zixiang Xu

REVIEW 5 major objections 5 minor 42 references

MTFusion: Reconstructing Any 3D Object from Single Image Using Multi-word Textual Inversion

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 MTFusion reconstructs a 3D object from a single image in under an hour.

desk verdict A credible engineering contribution whose quantitative claims outrun the disclosed evaluation protocol. read the letter →

arxiv 2411.12197 v1 pith:CGYWEQH7 submitted 2024-11-19 cs.CV cs.MM

Yu Liu , Ruowei Wang , Jiaqi Li , Zixiang Xu , Qijun Zhao This is my paper · ORCID

classification cs.CVcs.MM

keywords 3Dreconstructionsingle-imagetextualinversionscoredistillationsamplingFlexiCubessigneddistancefunctionhashgridencodinggradient-freeoptimization

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

MTFusion is a two-stage pipeline that takes one photograph of an object and outputs a textured 3D mesh. The first stage learns a multi-word pseudo-prompt, with separate learned tokens for style, object type, and residual detail, using a gradient-free evolutionary optimizer rather than the single-word textual inversion used by earlier methods. The second stage uses that prompt plus the input image to drive score-distillation training of a FlexiCubes mesh whose parameters are produced by an SDF decoder network with hashgrid positional encoding. The authors report that the resulting meshes beat prior NeRF-based image-to-3D methods on PSNR, LPIPS, and CLIP similarity, and that the full reconstruction takes about 55 minutes instead of 90.

What carries the argument

The central machinery is multi-word textual inversion: a prompt template of the form 'a <style> image of <object> <etc>', where <style>, <object>, and <etc> are pseudo-tokens initialized by CLIP-similarity weighting and optimized with CMA-ES in a PCA-reduced subspace, without backpropagating through the diffusion model. The second piece is enhanced FlexiCubes: instead of optimizing FlexiCubes weights directly, an MLP with multi-resolution hashgrid positional encoding outputs SDF values, vertex offsets, and the per-cube weights, enabling faster convergence and finer surface detail under score distillation. The two pieces are coupled by training the 3D model with both an SDS loss from the pseudo-prompt and a reconstruction loss at the reference view.

What would settle it

Render the output mesh at a held-out camera angle and compare it with a ground-truth photograph from that angle using image-similarity metrics; if the mesh visibly collapses, loses texture, or scores near the reference-view baseline, the claimed 3D fidelity is not real.

Watch

Extended reading notes

Core claim

The paper claims that representing a single input image by several learnable text embeddings instead of one makes SDS-based image-to-3D reconstruction more faithful, and that replacing a NeRF with FlexiCubes augmented by a hashgrid-encoded SDF decoder removes the surface artifacts typical of Marching Cubes extraction and speeds up training. At the core is a claim of joint benefit: the multi-word textual inversion supplies richer semantic conditioning (style, object, residual) that the diffusion prior can use, while the enhanced FlexiCubes representation keeps geometry and texture disentangled. The paper's numbers, PSNR 21.55, LPIPS 0.12, and CLIP-similarity 0.84 on the NeRF4 and RealFusion15 images, are offered as evidence that this joint design surpasses RealFusion, Make-It-3D, and Magic123, and that the full pipeline reconstructs a textured mesh in about 55 minutes.

Load-bearing premise

The load-bearing premise is that the reported image-similarity scores were computed at viewpoints not seen during training, so they truly measure 3D reconstruction quality rather than how well the model reproduces the single input photograph.

Editorial extensions

If this is right

If the reported numbers hold, single-image 3D reconstruction can be done with explicit meshes and a hybrid SDF representation, avoiding the Marching Cubes artifacts that NeRF-based baselines inherit.
Multi-word textual inversion gives the diffusion prior separate conditioning signals for style, object identity, and residual detail; ablations in the paper show that each added token improves PSNR/LPIPS over single-word embeddings.
The gradient-free evolutionary optimization removes the need to backpropagate through the diffusion model during prompt inversion, cutting the overall reconstruction time from roughly 90 to 55 minutes.
The hashgrid-encoded SDF decoder improves training stability and speed compared with vanilla FlexiCubes, which the paper demonstrates on text-only generation and in the full pipeline.

Reading between the lines

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

The paper's quantitative comparison rests on 19 test images and does not report a novel-view geometry metric; an editorial inference is that a larger benchmark with explicit multi-view evaluation would be needed to confirm that the PSNR/LPIPS gains reflect 3D coherence rather than reference-view fitting.
The multi-word inversion idea is representation-agnostic: the same style/object/residual token scheme could likely be paired with other differentiable surface representations, though the paper does not test this.
Because the texture network and geometry network are separate, MTFusion could be extended to edit or re-texture the reconstructed mesh by swapping the prompt tokens, which the paper does not explore.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

MTFusion is worth a look if you work on single-image 3D reconstruction. The actual novelty is limited — it merges multi-attribute textual inversion from [1], gradient-free CMA-ES from [4], and FlexiCubes with a hashgrid SDF decoder, similar in spirit to Fantasia3D. But the merge is done cleanly, and the two-stage pipeline makes sense: extract a richer pseudo-prompt than vanilla textual inversion, then use it to guide SDS with an explicit-surface representation. The ablation in Table 1 supports the design choices: adding style+object beats single-word, and the enhanced FlexiCubes beats vanilla. That is real evidence the components matter.

The soft spots are all in Section 4. Table 1 has no error bars, no per-dataset breakdown, and no statement of which camera views the metrics are computed over. Since the method is explicitly supervised with a reconstruction loss at the reference view, if PSNR/LPIPS/CLIP are measured at that same view, the comparison partly rewards fitting the input image rather than 3D generalization. The text claims elevated CLIP-similarity indicates 3D coherence, but CLIP-similarity at the reference view wouldn't show that. The 50% speedup is stated without any timing experiment. And there's no code release, so the numbers can't be checked.

The central idea — that multi-word gradient-free textual inversion plus FlexiCubes improves fidelity and training speed — is plausible and not oversold. I don't see a load-bearing flaw in the method itself; the problem is the evidence. The paper needs a disclosed evaluation protocol, per-scene results with error bars, and ideally a comparison at novel views to back the 3D coherence claim. If those are added, the claims are likely to hold up.

This deserves peer review rather than a desk reject: it's a coherent, honest system paper with useful empirical comparisons, and the evaluation gaps are fixable. I'd send it to a serious referee, with the instruction to focus on the metric protocol. For my own work I probably wouldn't cite it in the next year, but I'd read the revised version.

Referee Report

5 major / 5 minor

Summary. The paper proposes MTFusion, a two-stage method for reconstructing a 3D textured mesh from a single image. In the first stage, a multi-word pseudo-prompt containing style, object, and residual tokens is optimized in the text-embedding space of Stable Diffusion using a gradient-free CMA-ES strategy. In the second stage, a FlexiCubes-based representation is trained with SDS loss at novel views and a 2D reconstruction loss at the reference view; an MLP with multi-resolution hashgrid positional encoding predicts SDF values and FlexiCubes parameters. Experiments compare MTFusion with RealFusion, Make-It-3D, and Magic123 on the NeRF4 and RealFusion15 datasets, reporting PSNR 21.55, LPIPS 0.12, and CLIP-similarity 0.84, as well as a reduction in reconstruction time from about 90 to 55 minutes.

Significance. If the quantitative claims are substantiated, MTFusion would be a practically useful baseline for single-image 3D reconstruction: it directly outputs a textured mesh, avoids the NeRF-to-mesh postprocessing used by several baselines, and demonstrates that gradient-free prompt optimization can be integrated into SDS-based 3D generation. The use of public baselines and Stable Diffusion v1.5 is a strength, and the ablation study attempts to isolate the contributions of the proposed components. However, the central claim of superiority over prior methods is currently not statistically or protocol-wise established, so the paper's contribution is better viewed as a promising direction than as a fully validated method.

major comments (5)

[Section 4.3, Table 1] The quantitative comparison does not state which camera views are used to compute PSNR, LPIPS, and CLIP-similarity. Since Section 3.3 explicitly applies a 2D reconstruction loss at the reference view for both geometry and appearance, evaluating at that view would partly reward fitting the input image rather than 3D generalization. Please disclose the view protocol and report novel-view metrics (e.g., PSNR and LPIPS on held-out views) to support the claim of superior 3D reconstruction.
[Section 4.3, Table 1] The evaluation is aggregated over only 19 images (NeRF4 plus RealFusion15) with no error bars, per-dataset breakdown, or significance tests. The claim that MTFusion 'surpasses' baselines is therefore not statistically supported; please add per-scene results, standard deviations, and at least paired significance tests.
[Section 4.3] The speedup claim is internally inconsistent: reducing reconstruction time from about 90 minutes to 55 minutes is a 39% reduction, not the stated 50%, and no timing protocol is described. Please either substantiate the claim with measured wall-clock times or iteration counts, or remove the quantitative speedup statement.
[Section 3.2 and Section 4.1] The proposed gradient-free optimization is not fully specified: the PCA dimension d for the projection matrix Wp, the CMA-ES population size and iteration count, the fixed noise timestep t used in evaluations, and the number of training iterations for the geometry and texture stages are all omitted. Without these details, the method cannot be reproduced or compared fairly.
[Section 4.3] The statement that an elevated CLIP-similarity score 'suggests a remarkable degree of 3D coherence between the reconstructed object and the reference view' is not supported, because CLIP-similarity computed at the reference view would measure 2D semantic alignment rather than multi-view consistency. A separate novel-view evaluation is needed for this interpretation.

minor comments (5)

[Section 3.2] The heading contains a typo ('W ord') and Section 3.3 contains 'geomotry'; please proofread the manuscript thoroughly.
[Section 4.3, Table 1] The row descriptions in the text do not match the table structure: 'Rows 4-6 show the ablation results on multi-word textual inversion' and 'the last two rows show the ablation results on the enhanced FlexiCubes representation' need to be aligned with the actual rows of Table 1.
[Section 4.1 and Section 4.3] Section 4.1 states that the entire reconstruction process takes approximately one hour, which is difficult to reconcile with the 55-minute figure in Section 4.3; please make the timing statements consistent.
[Section 3.2] The novelty of 'multi-word textual inversion' should be clarified relative to reference [1] (multi-attribute inversion) and reference [4] (gradient-free textual inversion); as written, the contribution overlaps substantially with both.
[Section 4.1] Please clarify whether code and data will be released, and specify the exact prompt template and Stable Diffusion version used in the experiments.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity; the method is benchmarked against external baselines, though reference-view metric ambiguity is a reporting concern.

full rationale

MTFusion's derivation is not circular in the sense defined here. Stage 1 fits a pseudo-prompt to the input image via a diffusion loss; Stage 2 conditions a FlexiCubes model on both that prompt and the image, using a reference-view reconstruction loss plus SDS loss at novel views. This is a standard personalization-plus-distillation pipeline: the fitted text embedding is an input representation, not a quantity that is simultaneously the predicted output. The quantitative comparison (Table 1) is against external baselines (RealFusion, Make-It-3D, Magic123), so the headline superiority claim is independently benchmarked rather than derived from self-citation. Self-citations (refs 7, 34, 35, 39) appear only as related-work context and are not load-bearing. The main concern—that PSNR/LPIPS/CLIP may be evaluated at the reference view, which is directly supervised—is a correctness/reporting ambiguity (the paper does not state the view protocol), not a reducibility-by-construction step; without knowing the protocol we cannot exhibit Eq. X = Eq. Y by construction. Thus no specific circular step rises to the required evidentiary standard.

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

The paper introduces no physical entities. The load-bearing assumptions are that Stable Diffusion priors generalize to arbitrary objects, that three pseudo-tokens capture the needed image attributes, and that stage-1 embeddings transfer to SDS. Key hyperparameters are unstated, making the method difficult to reconstruct.

free parameters (5)

PCA embedding dimension d for the projection matrix Wp = not reported
The size of the subspace searched by CMA-ES controls the expressiveness of pseudo-token embeddings; the paper never states d.
CMA-ES population size and number of iterations = not reported
These determine whether the textual inversion stage converges to a useful embedding; no values are given.
Fixed noise timestep t used during textual inversion evaluations = not reported
The method fixes t during each evaluation, but the chosen value is not stated; it affects the diffusion loss landscape.
Grid resolution for FlexiCubes = 128
Chosen by hand as a speed-quality tradeoff; geometry detail depends on it.
Training iteration counts = 5000 coarse, 5000 fine, 500 warm-up
The 3D stage uses coarse-to-fine training with these budgets; the paper does not ablate them.

assumptions (5)

domain assumption The pretrained Stable Diffusion v1.5 model, used through SDS, provides a sufficient 3D prior for arbitrary objects.
Used in Section 3.3 to supervise geometry and texture at novel views; no analysis of coverage for unseen object classes.
ad hoc to paper A prompt template of three pseudo-tokens (style, object, residual) can capture the information needed for 3D reconstruction.
Introduced in Section 3.2; the sufficiency of this decomposition is asserted, not demonstrated.
domain assumption CMA-ES optimization in a PCA subspace, with a fixed noise level, yields an embedding that transfers to SDS-based mesh generation.
The stage-1 objective (diffusion denoising loss on the image) is optimized separately from the stage-2 SDS objective; transfer is assumed.
domain assumption Differentiable rendering through FlexiCubes, with the SDF decoder, provides unbiased gradients for SDS optimization.
Borrowed from [32] and [3]; no verification is included in this paper.
domain assumption The combination of reference-view reconstruction loss and SDS loss at novel views is a sufficient objective for high-fidelity reconstruction.
Used throughout Section 3.3; the paper does not study the tradeoff between the two loss terms.

how reviews work

0 comments

Cite this review

Pith. "Pith review of MTFusion: Reconstructing Any 3D Object from Single Image Using Multi-word Textual Inversion." pith.science (2026). https://pith.science/paper/CGYWEQH7

@misc{pith2026241112197,
  author       = {Pith},
  title        = {Pith review of: MTFusion: Reconstructing Any 3D Object from Single Image Using Multi-word Textual Inversion},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CGYWEQH7}},
  note         = {Machine review of arXiv:2411.12197}
}

read the original abstract

Reconstructing 3D models from single-view images is a long-standing problem in computer vision. The latest advances for single-image 3D reconstruction extract a textual description from the input image and further utilize it to synthesize 3D models. However, existing methods focus on capturing a single key attribute of the image (e.g., object type, artistic style) and fail to consider the multi-perspective information required for accurate 3D reconstruction, such as object shape and material properties. Besides, the reliance on Neural Radiance Fields hinders their ability to reconstruct intricate surfaces and texture details. In this work, we propose MTFusion, which leverages both image data and textual descriptions for high-fidelity 3D reconstruction. Our approach consists of two stages. First, we adopt a novel multi-word textual inversion technique to extract a detailed text description capturing the image's characteristics. Then, we use this description and the image to generate a 3D model with FlexiCubes. Additionally, MTFusion enhances FlexiCubes by employing a special decoder network for Signed Distance Functions, leading to faster training and finer surface representation. Extensive evaluations demonstrate that our MTFusion surpasses existing image-to-3D methods on a wide range of synthetic and real-world images. Furthermore, the ablation study proves the effectiveness of our network designs.

Figures

Figures reproduced from arXiv: 2411.12197 by the authors.

**Figure 1.** Given a single-view image, MTFusion generates a textured mesh using supervision from the image and a pseudo-prompt. This approach leads to a naturally occurring question: can text-to-3D generation be extended to image-to-3D generation by mapping the input image to the text embedding space? The key point of this strategy lies in finding a text embedding that could faithfully reconstruct the given image. Some attem… view at source ↗

**Figure 2.** Overview of MTFusion. Our approach extracts a textual description from the input image and constrains the 3D model based on this description and the image. 3.1 Preliminary Latent Diffusion Models. LDMs [29] are machine learning models designed to learn the underlying structure of a dataset by mapping it to a lower-dimensional latent space. It involves two fundamental processes: a forward diffusion process that progr… view at source ↗

**Figure 3.** Overview of our proposed Multi-Word Textual Inversion. The optimization of text embedding is based on a gradient-free approach, which iteratively employs an evolution strategy to explore and exploit pseudo-token embeddings. this optimization process, all other parts of text embeddings within the prompt and model parameters remain fixed. Different from the vanilla textual inversion, which sets the prompt template as … view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Qualitative comparison with RealFusion, Make-It-3D and Magic123 on synthetic and real-world images. Note that for real-world images, we first remove the background with [26], then use the preprocessed RGB-A images for 3D modeling. Due to the utilization of Multi-Word …

**Figure 5.** Figure 5: Ablation study on the enhanced FlexiCubes with hashgrid positional encoding. During the 3D mesh generation process from a given textual description (”A pineapple.”), our enhanced FlexiCubes shows better training stability and robustness [PITH_FULL_IMAGE:figures/full_…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 34 canonical work pages

[1]

arXiv preprint arXiv:2311.11919 (2023)

Agarwal, A., Karanam, S., Shukla, T., Srinivasan, B.V.: An image is worth multi- ple words: Multi-attribute inversion for constrained text-to-image synthesis. arXiv preprint arXiv:2311.11919 (2023)

arXiv 2023
[4]

In: ACMMM (2023)

Fei, Z., Fan, M., Huang, J.: Gradient-free textual inversion. In: ACMMM (2023)

work page 2023
[2]

In: CVPR (2021)

Barron, J.T., Mildenhall, B., Tancik, M., Hedman, P., Martin-Brualla, R., Srini- vasan, P.P.: Mip-nerf: A multiscale representation for anti-aliasing neural radiance fields. In: CVPR (2021)

work page 2021
[3]

In: ICCV (2023)

Chen, R., Chen, Y., Jiao, N., Jia, K.: Fantasia3D: Disentangling geometry and appearance for high-quality text-to-3D content creation. In: ICCV (2023)

work page 2023
[5]

In: ICLR (2023)

Gal, R., Alaluf, Y., Atzmon, Y., Patashnik, O., Bermano, A.H., Chechik, G., Cohen-Or, D.: An image is worth one word: Personalizing text-to-image gener- ation using textual inversion. In: ICLR (2023)

work page 2023
[6]

NeurIPS (2022)

Gao, J., Shen, T., Wang, Z., Chen, W., Yin, K., Li, D., Litany, O., Gojcic, Z., Fidler, S.: Get3d: A generative model of high quality 3d textured shapes learned from images. NeurIPS (2022)

work page 2022
[7]

Journal of Image and Graphics 28(5), 1434–1444 (2023)

Gaoli, S., Shudi, X., Qijun, Z.: Soft threshold denoising and video data fusion- relevant low-quality 3d face recognition. Journal of Image and Graphics 28(5), 1434–1444 (2023)

work page 2023
[8]

Evolutionary computation 11(1), 1–18 (2003)

Hansen, N., M¨ uller, S.D., Koumoutsakos, P.: Reducing the time complexity of the derandomized evolution strategy with covariance matrix adaptation (cma-es). Evolutionary computation 11(1), 1–18 (2003)

work page 2003

Show all 42 references

[9]

ISPRS (1998)

Van den Heuvel, F.A.: 3D reconstruction from a single image using geometric constraints. ISPRS (1998)

1998
[10]

In: NeurIPS (2020)

Ho, J., Jain, A., Abbeel, P.: Denoising diffusion probabilistic models. In: NeurIPS (2020)

2020
[11]

In: TOG (2002)

Ju, T., Losasso, F., Schaefer, S., Warren, J.: Dual contouring of hermite data. In: TOG (2002)

2002
[12]

In: ICOSS (2013)

Khilar, R., Chitrakala, S., SelvamParvathy, S.: 3D image reconstruction: Tech- niques, applications and challenges. In: ICOSS (2013)

2013
[13]

JCH (2014)

Koutsoudis, A., Vidmar, B., Ioannakis, G., Arnaoutoglou, F., Pavlidis, G., Chamzas, C.: Multi-image 3D reconstruction data evaluation. JCH (2014)

2014
[14]

In: ICLR (2023)

Kwon, M., Jeong, J., Uh, Y.: Diffusion models already have a semantic latent space. In: ICLR (2023)

2023
[15]

TOG (2020)

Laine, S., Hellsten, J., Karras, T., Seol, Y., Lehtinen, J., Aila, T.: Modular primi- tives for high-performance differentiable rendering. TOG (2020)

2020
[16]

In: CVPR (2018)

Liao, Y., Donne, S., Geiger, A.: Deep marching cubes: Learning explicit surface representations. In: CVPR (2018)

2018
[17]

In: CVPR (2023)

Lin, C.H., Gao, J., Tang, L., Takikawa, T., Zeng, X., Huang, X., Kreis, K., Fidler, S., Liu, M.Y., Lin, T.Y.: Magic3D: High-resolution text-to-3D content creation. In: CVPR (2023)

2023
[18]

In: SIGGRAPH (1987)

Lorensen, W.E., Cline, H.E.: Marching cubes: A high resolution 3D surface con- struction algorithm. In: SIGGRAPH (1987)

1987
[19]

In: CVPR (2023)

Melas-Kyriazi, L., Laina, I., Rupprecht, C., Vedaldi, A.: Realfusion: 360deg recon- struction of any object from a single image. In: CVPR (2023)

2023
[20]

In: ECCV (2020) MTFusion 15

Mildenhall, B., Srinivasan, P.P., Tancik, M., Barron, J.T., Ramamoorthi, R., Ng, R.: Nerf: Representing scenes as neural radiance fields for view synthesis. In: ECCV (2020) MTFusion 15

2020
[21]

TOG (2022)

M¨ uller, T., Evans, A., Schied, C., Keller, A.: Instant neural graphics primitives with a multiresolution hash encoding. TOG (2022)

2022
[22]

In: CVPR (2022)

Munkberg, J., Hasselgren, J., Shen, T., Gao, J., Chen, W., Evans, A., M¨ uller, T., Fidler, S.: Extracting triangular 3d models, materials, and lighting from images. In: CVPR (2022)

2022
[23]

In: IEEE visualization 2004 (2004)

Nielson, G.M.: Dual marching cubes. In: IEEE visualization 2004 (2004)

2004
[24]

In: ICLR (2023)

Poole, B., Jain, A., Barron, J.T., Mildenhall, B.: Dreamfusion: Text-to-3D using 2D diffusion. In: ICLR (2023)

2023
[25]

In: ICLR (2024)

Qian, G., Mai, J., Hamdi, A., Ren, J., Siarohin, A., Li, B., Lee, H.Y., Skorokhodov, I., Wonka, P., Tulyakov, S., Ghanem, B.: Magic123: One image to high-quality 3D object generation using both 2D and 3D diffusion priors. In: ICLR (2024)

2024
[26]

Pattern Recognition 106, 107404 (2020)

Qin, X., Zhang, Z., Huang, C., Dehghan, M., Zaiane, O.R., Jagersand, M.: U2- net: Going deeper with nested u-structure for salient object detection. Pattern Recognition 106, 107404 (2020)

2020
[27]

In: ICML (2021)

Radford, A., Kim, J.W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., et al.: Learning transferable visual models from natural language supervision. In: ICML (2021)

2021
[28]

In: Optimization: Methods and Applications, Possibilities and Limitations (1989)

Rechenberg, I.: Evolution strategy: Nature’s way of optimization. In: Optimization: Methods and Applications, Possibilities and Limitations (1989)

1989
[29]

In: CVPR (2022)

Rombach, R., Blattmann, A., Lorenz, D., Esser, P., Ommer, B.: High-resolution image synthesis with latent diffusion models. In: CVPR (2022)

2022
[30]

NeurIPS (2022)

Schuhmann, C., Beaumont, R., Vencu, R., Gordon, C., Wightman, R., Cherti, M., Coombes, T., Katta, A., Mullis, C., Wortsman, M., et al.: Laion-5b: An open large-scale dataset for training next generation image-text models. NeurIPS (2022)

2022
[31]

In: NeurIPS (2021)

Shen, T., Gao, J., Yin, K., Liu, M.Y., Fidler, S.: Deep marching tetrahedra: a hybrid representation for high-resolution 3D shape synthesis. In: NeurIPS (2021)

2021
[32]

TOG (2023)

Shen, T., Munkberg, J., Hasselgren, J., Yin, K., Wang, Z., Chen, W., Gojcic, Z., Fidler, S., Sharp, N., Gao, J.: Flexible isosurface extraction for gradient-based mesh optimization. TOG (2023)

2023
[33]

In: ICLR (2021)

Song, J., Meng, C., Ermon, S.: Denoising diffusion implicit models. In: ICLR (2021)

2021
[34]

In: ICANN (2023)

Su, P., Zhao, Q., Pan, F., Gao, F.: Cascaded network-based single-view bird 3d reconstruction. In: ICANN (2023)

2023
[35]

In: ACCV (2020)

Sun, S., Zhu, Z., Dai, X., Zhao, Q., Li, J.: Weakly-supervised reconstruction of 3d objects with large shape variation from single in-the-wild images. In: ACCV (2020)

2020
[36]

In: ICLR (2024)

Tang, J., Ren, J., Zhou, H., Liu, Z., Zeng, G.: Dreamgaussian: Generative gaussian splatting for efficient 3D content creation. In: ICLR (2024)

2024
[37]

In: ICCV (2023)

Tang, J., Wang, T., Zhang, B., Zhang, T., Yi, R., Ma, L., Chen, D.: Make-it-3D: High-fidelity 3D creation from a single image with diffusion prior. In: ICCV (2023)

2023
[38]

In: CVPR (2023)

Wang, H., Du, X., Li, J., Yeh, R.A., Shakhnarovich, G.: Score jacobian chaining: Lifting pretrained 2D diffusion models for 3D generation. In: CVPR (2023)

2023
[39]

In: CVPR (2023)

Wang, R., Liu, Y., Su, P., Zhang, J., Zhao, Q.: 3d semantic subspace traverser: Empowering 3d generative model with shape editing capability. In: CVPR (2023)

2023
[40]

In: NeurIPS (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. In: NeurIPS (2024)

2024
[41]

In: CVPR (2018)

Zhang, R., Isola, P., Efros, A.A., Shechtman, E., Wang, O.: The unreasonable effectiveness of deep features as a perceptual metric. In: CVPR (2018)

2018
[42]

arXiv preprint arXiv:2312.14216 (2023)

Zhao, B.N., Xiao, Y., Xu, J., Jiang, X., Yang, Y., Li, D., Itti, L., Vineet, V., Ge, Y.: Dreamdistribution: Prompt distribution learning for text-to-image diffusion models. arXiv preprint arXiv:2312.14216 (2023)

2023 arXiv

Pith tools

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

[1] [1]

arXiv preprint arXiv:2311.11919 (2023)

Agarwal, A., Karanam, S., Shukla, T., Srinivasan, B.V.: An image is worth multi- ple words: Multi-attribute inversion for constrained text-to-image synthesis. arXiv preprint arXiv:2311.11919 (2023)

arXiv 2023

[2] [4]

In: ACMMM (2023)

Fei, Z., Fan, M., Huang, J.: Gradient-free textual inversion. In: ACMMM (2023)

work page 2023

[3] [2]

In: CVPR (2021)

Barron, J.T., Mildenhall, B., Tancik, M., Hedman, P., Martin-Brualla, R., Srini- vasan, P.P.: Mip-nerf: A multiscale representation for anti-aliasing neural radiance fields. In: CVPR (2021)

work page 2021

[4] [3]

In: ICCV (2023)

Chen, R., Chen, Y., Jiao, N., Jia, K.: Fantasia3D: Disentangling geometry and appearance for high-quality text-to-3D content creation. In: ICCV (2023)

work page 2023

[5] [5]

In: ICLR (2023)

Gal, R., Alaluf, Y., Atzmon, Y., Patashnik, O., Bermano, A.H., Chechik, G., Cohen-Or, D.: An image is worth one word: Personalizing text-to-image gener- ation using textual inversion. In: ICLR (2023)

work page 2023

[6] [6]

NeurIPS (2022)

Gao, J., Shen, T., Wang, Z., Chen, W., Yin, K., Li, D., Litany, O., Gojcic, Z., Fidler, S.: Get3d: A generative model of high quality 3d textured shapes learned from images. NeurIPS (2022)

work page 2022

[7] [7]

Journal of Image and Graphics 28(5), 1434–1444 (2023)

Gaoli, S., Shudi, X., Qijun, Z.: Soft threshold denoising and video data fusion- relevant low-quality 3d face recognition. Journal of Image and Graphics 28(5), 1434–1444 (2023)

work page 2023

[8] [8]

Evolutionary computation 11(1), 1–18 (2003)

Hansen, N., M¨ uller, S.D., Koumoutsakos, P.: Reducing the time complexity of the derandomized evolution strategy with covariance matrix adaptation (cma-es). Evolutionary computation 11(1), 1–18 (2003)

work page 2003