Pith. sign in

REVIEW 4 major objections 6 minor 68 references

SVGDreamer++: Advancing Editability and Diversity in Text-Guided SVG Generation

T0 review · 4 major / 6 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 Hierarchical masks make generated SVGs editable and diverse.

desk verdict SVGDreamer++ adds two genuinely new components and a coherent pipeline, but its state-of-the-art claim rests on an evaluation protocol that is never defined and may reward copying the supervision image rather than text-to-SVG generation. read the letter →

arxiv 2411.17832 v2 pith:UDRWE5RK submitted 2024-11-26 cs.CV cs.AI

Ximing Xing , Qian Yu , Chuang Wang , Haitao Zhou , Jing Zhang , Dong Xu This is my paper · ORCID
classification cs.CVcs.AI
keywords vectorgraphicsSVGgenerationtext-to-SVGhierarchicalvectorizationsegmentation-guidedoptimizationscoredistillationeditablediverse
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

The paper aims to show that text-guided SVG generation can produce outputs that are not only visually competitive with raster text-to-image results but also genuinely editable, meaning a designer can select, recolor, recombine, or replace individual objects and object parts. To do this, it introduces a hierarchical vectorization step (HIVE) that uses image-segmentation masks to supervise separate groups of Bézier paths, and a particle-based score distillation (VPSD) that treats several sets of SVG parameters as samples from a distribution. It also changes the number of paths during optimization, splitting or cloning paths in regions where the loss gradient suggests under- or over-representation. The paper reports that the combined method outperforms prior text-to-SVG systems on all six evaluated metrics and supports six vector styles.

What carries the argument

The load-bearing mechanism is the hierarchical mask loss $L_{HIVE} = \sum_i \|\hat{M}_i \odot I - \hat{M}_i \odot x\|_2^2 + \sum_i\sum_j \|\tilde{M}^j_i \odot I_i - \tilde{M}^j_i \odot x_i\|_2^2$, where $\hat{M}_i$ are object masks and $\tilde{M}^j_i$ are part masks. The masks come from a two-stage process: coarse masks from a text-grounded segmentation, then fine masks conditioned on the current control-point coordinates. Because the loss is masked per region, each path group is supervised only inside its own semantic region, which is what enforces decoupling and editability. The adaptive primitive control (Algorithm 1) is the companion mechanism: it prunes near-transparent paths, splits over-represented large paths, and clones under-represented small paths, using the gradient magnitude of the HIVE loss to locate regions that need more primitives.

What would settle it

Run the full SVGDreamer++ pipeline on a prompt set with fine-grained texture such as fur, hair, or foliage, and compare the variant with adaptive primitive control disabled against the variant with it enabled while holding all other hyperparameters fixed; also vary the gradient threshold $\tau_c$ across a range. If FID and PSNR do not improve on the high-texture set, or if the threshold sensitivity is so sharp that small changes reverse the gain, the adaptive-control claim is not robust. A simpler specific check is to render the gradient map of the HIVE loss for a target whose geometry is already perfectly covered but whose color is wrong; if the algorithm still splits paths, the heuristic is misfiring.

Watch

Extended reading notes

Core claim

SVGDreamer++ proposes that the right supervision signal for text-to-SVG is not the full raster rendering but a hierarchy of masks: object-level masks obtained by grounding each noun in the prompt, and part-level masks obtained by conditioning a segmentation model on the current control-point positions of each object. With these masks as per-region losses, vector paths stay inside their semantic region, so the final SVG can be edited at object and part level. The same optimization is driven by VPSD, which maintains k particles (sets of control points and colors), estimates their distribution with a LoRA-adapted diffusion model, and reweights particles by a pretrained aesthetic reward, yielding diverse, less saturated images with faster convergence. Combined with adaptive primitive control, which splits large paths that cover too much geometry and clones small paths that cover too little based on the gradient magnitude of the HIVE loss, the result is a text-to-SVG pipeline that the authors claim achieves the best scores on FID, PSNR, CLIP, BLIP, Aesthetic, and HPS among the methods they compare.

Load-bearing premise

The method assumes that a large gradient magnitude in the HIVE loss marks a spot where a path should be split or cloned, rather than a spot that needs a different color, opacity, or a corrected boundary; if that mapping is unreliable, added paths land in the wrong places and the reported quality gain from adaptive primitive control could disappear.

Editorial extensions

If this is right

  • If HIVE works as claimed, vector objects generated from multi-noun prompts can be decoupled into independently editable components, enabling recombination of foregrounds and backgrounds from different generated SVGs.
  • Part-level masks would allow local edits, such as swapping a character's cloak or weapon, without re-running generation or manually tracing paths.
  • VPSD's particle distribution would yield multiple SVG samples for one prompt with more diversity and less color over-saturation than SDS-based methods, at roughly half the optimization time when reward feedback is used.
  • Adaptive primitive control would remove the need to choose a fixed path count in advance, improving detail where needed while keeping optimization faster than a uniformly large path budget.
  • With the six primitive-style settings (icon, pixel art, low-poly, painting, sketch, ink-wash), the same pipeline can serve as a vector-asset generator for posters, stickers, and item art.

Reading between the lines

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

  • The gradient-magnitude heuristic for path splitting and cloning is really a proxy for reconstruction error; one could test it directly by comparing against a policy that adds paths based on masked reconstruction error or segmentation uncertainty, to see whether gradient magnitude is the right trigger.
  • The hierarchy idea could be pushed one level further: if masks are available, the same per-region loss could be used to learn a reusable compositional prior, such as a library of independently editable object prototypes, rather than optimizing each prompt from scratch.
  • Because the method's editability ceiling is set by the text-to-image model's ability to separate concepts in its attention, improvements in text-to-image grounding should transfer automatically; the paper itself notes this dependence in its limitations.
  • A natural stress test is prompts with several overlapping same-category objects, such as two astronauts, where attention-based initialization and part masks may conflict; the paper shows one such case, but systematic scaling of the number of objects would reveal where the decomposition degrades.
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

4 major / 6 minor

Summary. The paper presents SVGDreamer++, a text-guided SVG generation method that extends the authors' earlier SVGDreamer with two new components: a Hierarchical Image VEctorization (HIVE) strategy that uses Grounded-SAM and SAM masks to supervise vectorization at both object and part levels, and an Adaptive Vector Primitives Control algorithm that dynamically splits, clones, or prunes vector paths during optimization based on gradient magnitudes. The method combines these components with the previously proposed Vectorized Particle-based Score Distillation (VPSD) and Reward Feedback Learning to improve visual quality, diversity, and editability. The authors claim state-of-the-art performance on six metrics, demonstrate object- and part-level editing, and show applications in poster design and vector asset generation.

Significance. If the central claims hold, the paper makes a useful contribution to text-to-SVG generation by addressing two practically important shortcomings: the lack of editability in the output SVGs and the difficulty of setting the number of vector primitives in advance. The HIVE framework's use of segmentation priors for part-level decomposition is a plausible and well-motivated idea, and the adaptive primitive control is a novel mechanism for controlling SVG complexity during optimization. The paper is also commendable for including qualitative comparisons of three vectorization approaches, ablations of particle counts, and a stated intention to release code. However, the main quantitative claim is currently supported by an incompletely specified evaluation protocol, and the contributions of the two new modules are not isolated by quantitative ablations; these issues need to be resolved before the claims can be fully accepted.

major comments (4)
  1. [Section 6, Table 1] The reference set for FID and the reference image for PSNR are never defined. The paper states that HIVE optimizes the SVG to match a raster image I generated by a latent diffusion model (Eq. 8), and the evaluation section (Section 6, 'Evaluation Metrics') only names the metrics without specifying the reference distribution or reference image. If the FID reference set and the PSNR reference are the LDM-generated images used internally by each method, then methods that receive a direct raster target (VectorFusion, SVGDreamer, SVGDreamer++) are structurally favored, while CLIP-based methods that do not use such a target are disadvantaged. This is load-bearing because the headline claim of 'best performance across all evaluated metrics' rests entirely on Table 1. Please specify the reference set, report the number of prompts and seeds, and provide the protocol so that the comparison is method-independent.
  2. [Section 4.2, Algorithm 1] The adaptive vector primitive control is central to the claimed visual quality improvement, but its effectiveness is supported only by qualitative visualizations (Figs. 11, 12). There is no quantitative ablation in Table 1 or elsewhere that isolates the contribution of this module, and no sensitivity analysis for the hand-tuned thresholds tau_opacity, tau_c, and tau_a. Since the algorithm's decision rule assumes that high gradient magnitude in the HIVE loss reliably indicates regions needing more paths, an ablation that compares the full method with and without adaptive control, along with a sensitivity study for the thresholds, is necessary to validate this load-bearing assumption.
  3. [Section 6.2.5, Table 2] Table 2 reports iteration counts and wall-clock times for VPSD with and without Reward Feedback Learning, but it does not report any quality metric for these two configurations. The text states that ReFL 'improves the aesthetic score of the SVG' and refers to Table 1, but Table 1 compares full methods rather than a controlled with/without-ReFL ablation. Please provide a quantitative ablation (e.g., CLIP score, aesthetic score, or HPS) for the same prompt set with and without ReFL, otherwise the aesthetic-improvement claim is not supported by the presented evidence.
  4. [Section 6, Dataset and Evaluation Metrics] The paper describes three prompt sets (single object, single object with surroundings, multiple objects) but does not state how many prompts are in each set, how many optimization runs per prompt are used, or whether Table 1 aggregates over all three sets. No error bars or variances are reported for any of the six metrics. Given that the reported differences between some methods are moderate (e.g., CLIP score 0.3001 vs. 0.3093), the absence of any statistical variability measure makes it difficult to assess whether the differences are meaningful. Please provide the number of prompts, seeds, and per-metric variances or confidence intervals.
minor comments (6)
  1. [Throughout] There are several typos and inconsistencies that should be corrected: 'SVGDremaer++' in Figure 1, 'purning' in Algorithm 1, 'editied' in Section 6.1.3, 'VPSD v.s. LSDS v.s. ASDS' in Section 6.2.3, and 'Low-ploy' in Figure 15.
  2. [Section 3.1.1, Eq. 1] The definition of the foreground attention maps M_FG is unclear: the softmax is written as softmax(Q K_i^T)/sqrt(d), but it is not specified over which dimension the softmax is applied, and the threshold used later to convert attention maps into binary masks is not stated.
  3. [Section 3.2, Eq. 6] The reward feedback loss in Eq. 6 is written as an expectation over y of psi(r(y, g_phi_est(y))), but the text describes a reweighting of samples collected from the LoRA model. The connection between the expectation form and the 'reweighted samples' description should be made explicit, and the role of the ReLU map psi should be clarified.
  4. [Algorithm 1] The operations RemovePath, SplitPath, and ClonePath are not formally defined. In particular, it is not specified how a path is split into two paths (whether control points are duplicated, whether colors are inherited, and how the new paths are initialized) or how a cloned path is placed relative to the original.
  5. [Section 6.2.4, Figure 13] The text says 'we limit our testing to a maximum of 6 particles,' but Figure 13 shows results for 1, 4, 8, and 16 particles, which exceeds this maximum. This inconsistency should be resolved, either by correcting the text or by explaining how the 16-particle results were produced.
  6. [Section 6.3.1, Figure 16] The poster-design comparison with raster T2I methods is presented only qualitatively. Since the paper emphasizes the advantages of vector posters in text clarity and editability, a small quantitative or user-study evaluation would strengthen this section, but this is not essential to the main claims.

Circularity Check

1 steps flagged · score 4.0 of 10

Technical derivation is largely self-contained; the clearest circularity is evaluative, not derivational: Table 1's PSNR has no stated reference image, and the only target image defined in the method is the LDM raster that Eq. (8) optimizes the SVG to match.

  1. fitted input called prediction [Sec. 6 'Evaluation Metrics' and Table 1, with Sec. 4.1 Eq. (8)]
    "LHIVE = Σ_i ( ˆMi ⊙ I − ˆMi ⊙ x)^2 + Σ_i Σ_j ( ˜Mj_i ⊙ I_i − ˜Mj_i ⊙ x_i)^2 (8) where I and I_i are the target image and the i-th object... Table 1 reports 'SVGDreamer++ 22.13 15.80 0.3093 0.4701 5.6124 0.2760' under 'FID ↓ PSNR ↑ CLIP Score ↑ BLIP Score ↑ Aesthetic ↑ HPS ↑'."

    Eq. (8) is the training objective: the rendered SVG x is driven pixel-wise toward a target raster I. The paper never states what the PSNR reference image is for Table 1; the only target image defined in the method is I, which in practice is an LDM-generated raster (Sec. 6.2.2: 'we input text prompt ... into the Latent Diffusion Model (LDM) [16] to generate raster images'). PSNR is a monotone transformation of MSE, so if the undeclared reference is I, the reported PSNR is the training loss relabeled as an evaluation metric. The same concern applies to FID if its reference set consists of LDM samples on the same prompts.

full rationale

Aside from the reference-image protocol gap, the paper's technical pipeline does not derive its contributions from self-defined targets. VPSD is an adaptation of variational score distillation (ProlificDreamer) with a LoRA estimator and an external ImageReward model; HIVE's masks come from Grounded-SAM/SAM; adaptive primitive control is a gradient-based heuristic. The self-citations to the authors' conference version [9] are load-bearing only as a baseline (SVGDreamer), but that prior work is published and not itself invoked as an unverified uniqueness theorem. The adaptive-control heuristic lacks sensitivity analysis, but that is a robustness concern, not circularity. The main circular step is evaluative: PSNR, and possibly FID, appear to measure reconstruction fidelity to the LDM image that the SVG was explicitly optimized to match. The Table 1 superiority claim should therefore be re-run against a disclosed, independent reference set before it is accepted as evidence of text-to-SVG superiority.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The method introduces no new physical or conceptual entities beyond the existing models it calls. The main burden rests on the external pretrained models (Stable Diffusion, SAM, LoRA, ImageReward) and on hand-set thresholds that control the adaptive path module.

free parameters (4)
  • opacity threshold (tau_opacity) = 0.05
    Used in Algorithm 1 to prune paths with near-zero opacity; no sensitivity analysis given.
  • gradient control threshold (tau_c) = 1e-5
    Determines when a path is split or cloned based on loss gradient magnitude; hand-set in Sec. 4.2.
  • area threshold (tau_a) = 20000 for canvas 1024, 10000 for canvas 768
    Distinguishes over-represented from under-represented paths; set per canvas size in Implementation Details.
  • reward feedback strength (lambda_r) = not stated
    Weights the reward loss in Eq. 7; value never specified in the paper.
assumptions (5)
  • domain assumption Pretrained Stable Diffusion provides a valid score function for text-conditioned image distributions
    Used throughout VPSD as the target distribution's noise predictor (Eq. 4).
  • domain assumption The LoRA network can approximate the distribution of rendered SVG parameters
    VPSD relies on a LoRA-finetuned network to estimate the score of rendered images (Eq. 4, 5).
  • domain assumption SAM and Grounded-SAM segment generated images accurately enough to guide vectorization
    HIVE uses SAM masks as supervision (Eq. 8); failure of segmentation would break editability and boundary quality.
  • ad hoc to paper High gradient magnitude in the HIVE loss indicates regions needing additional vector paths
    The adaptive control algorithm (Sec. 4.2) uses this correlation without a proof or sensitivity analysis.
  • standard math DiffVG renders SVGs with reliable gradients for optimization
    The whole optimization pipeline backpropagates through DiffVG, as in prior work.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SVGDreamer++: Advancing Editability and Diversity in Text-Guided SVG Generation." pith.science (2026). https://pith.science/paper/UDRWE5RK

@misc{pith2026241117832,
  author       = {Pith},
  title        = {Pith review of: SVGDreamer++: Advancing Editability and Diversity in Text-Guided SVG Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UDRWE5RK}},
  note         = {Machine review of arXiv:2411.17832}
}
read the original abstract

Recently, text-guided scalable vector graphics (SVG) synthesis has demonstrated significant potential in domains such as iconography and sketching. However, SVGs generated from existing Text-to-SVG methods often lack editability and exhibit deficiencies in visual quality and diversity. In this paper, we propose a novel text-guided vector graphics synthesis method to address these limitations. To enhance the editability of output SVGs, we introduce a Hierarchical Image VEctorization (HIVE) framework that operates at the semantic object level and supervises the optimization of components within the vector object. This approach facilitates the decoupling of vector graphics into distinct objects and component levels. Our proposed HIVE algorithm, informed by image segmentation priors, not only ensures a more precise representation of vector graphics but also enables fine-grained editing capabilities within vector objects. To improve the diversity of output SVGs, we present a Vectorized Particle-based Score Distillation (VPSD) approach. VPSD addresses over-saturation issues in existing methods and enhances sample diversity. A pre-trained reward model is incorporated to re-weight vector particles, improving aesthetic appeal and enabling faster convergence. Additionally, we design a novel adaptive vector primitives control strategy, which allows for the dynamic adjustment of the number of primitives, thereby enhancing the presentation of graphic details. Extensive experiments validate the effectiveness of the proposed method, demonstrating its superiority over baseline methods in terms of editability, visual quality, and diversity. We also show that our new method supports up to six distinct vector styles, capable of generating high-quality vector assets suitable for stylized vector design and poster design. Code and demo will be released at: http://ximinng.github.io/SVGDreamerV2Project/

Figures

Figures reproduced from arXiv: 2411.17832 by the authors.

Figure 1
Figure 1. SVGs produced by SVGDremaer++. Given a text prompt, SVGDreamer++ can generate a variety of vector graphics. SVGDreamer++ is a versatile tool that can work with various vector styles without being limited to a specific prompt suffix. We utilize various colored suffixes to indicate different styles. The style is governed by vector primitives [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The pipeline of SIVE. SIVE comprises two primary modules: primitive initialization and semantic-aware optimization. The primitive initialization module leverages diffusion model attention priors to initially delineate the paths of the corresponding vector objects. Subsequently, an attention-based mask loss function is introduced to facilitate the hierarchical optimization of these vector objects. 3 THE SVGDREAMER AP… view at source ↗
Figure 3
Figure 3. The process of Vectorized Particle-based Score Distillation. VPSD accepts k sets of SVG parameters as input. VPSD models SVG as a distribution of vector paths and color parameters, estimating these parameters through the application of the LoRA network. Through the estimation of the SVG parameter distribution, VPSD achieves a greater diversity of outputs compared to VF [7]. Moreover, to enhance the aesthetic quality… view at source ↗
Figures from the paper (12 more)
Figure 4
Figure 4. Figure 4: Overview of SVGDreamer++. Our method consists of two phases: Hierarchical image vectorization (Sec. 4.1) and optimized synthesis of diverse SVGs via VPSD (Sec. 3.2). And an additional module, called Adaptive Vector Primitives Control (Sec. 4.2), can be plugged into HIV…
Figure 5
Figure 5. Figure 5: The limitation of SIVE. When the cross attention map extracted from the LDM has a much lower resolution (e.g., 32x32) compared to the target vector graphic (e.g., 512x512), the results may have inaccurate boundaries. early iteration stage. This method saves 2 times the…
Figure 7
Figure 7. Figure 7: The HIVE loss gradient map tracing. The gradient map is derived by calculating the Jacobian matrix of the HIVE loss function and its gradient. Regions exhibiting higher gradient strengths suggest that the reconstruction is less effective, thereby requiring the addition…
Figure 6
Figure 6. Figure 6: Our Adaptive Vector Primitives Control scheme. Top row (Over-Represented): When a large graphic is used to represent small￾scale geometry, we address this by splitting the graphic into two new graphics, each exactly half the size of the original. Bottom row (Under￾Repr…
Figure 8
Figure 8. Figure 8: Qualitative comparison of SVGDreamer++ vs. the state-of-the-art Text-to-SVG methods. Note that DiffSketcher was originally designed for vector sketch generation; therefore, we re-implemented it to generate RGB vector images. SVGDreamer++ is capable of composing complex…
Figure 9
Figure 9. Figure 9: The editability of SVGDreamer++ results. Our process initiates with the examination of two SVGs generated by SVGDreamer++ (SVG1 and SVG2), where we first illustrate the decoupled vector elements at the object level (BG1, FG1, BG2, FG2 and FG3). Subsequently, we generat…
Figure 11
Figure 11. Figure 11: Comparison of HIVE vectorization process with LIVE and SIVE. HIVE outperforms LIVE in terms of accuracy for vector path con￾trol by strategically utilizing gradient strength to guide path adjustments and accurately determine control points for operations such as cloni…
Figure 12
Figure 12. Figure 12: Our Adaptive Vector Primitives Control behavior. To il￾lustrate the behavior of Adaptive Vector Primitives Control, we visual￾ize its vectorization process. Employing the same random initialization to generate 100 paths, we implement the Adaptive Vector Primitives Con…
Figure 14
Figure 14. Figure 14: Effects of the Reward Learning in VPSD. When employing Reward Learning, the visual quality of the generated results is signifi￾cantly enhanced. DreamFusion reconstructed 3D models. The main distinc￾tion between ASDS [8] and LSDS [7, 21] lies in the aug￾mentation of th…
Figure 15
Figure 15. Figure 15: SVG diversity generated by SVGDreamer++. We set the number of vector particles in SVGDreamer++ to 4 to synthesize diverse results. The results show that our method can maintain SVG quality and has variety. in the number of iterations by almost 50%, resulting in a 50% …
Figure 16
Figure 16. Figure 16: Qualitative comparison between SVGDreamer++ vector poster synthesis and state-of-the-art raster poster synthesis meth￾ods. The column on the left represents the input text prompt used to generate the poster and the font symbols in the poster. and graphic elements. Whi…
Figure 17
Figure 17. Figure 17: The vector assets generated by SVGDreamer++. We present a curated collection of vector assets encompassing four distinct styles: character portraits, graphic portraits, video game items, and vector stickers. Leveraging text descriptions, SVGDreamer++ can generate an e…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

68 extracted references · 56 canonical work pages

  1. [1]

    CLIPDraw: Exploring text-to-drawing synthesis through language- image encoders,

    K. Frans, L. Soros, and O. Witkowski, “CLIPDraw: Exploring text-to-drawing synthesis through language- image encoders,” in Advances in Neural Information Pro- cessing Systems (NeurIPS), 2022. 1, 4, 9, 10, 13

    work page 2022

  2. [2]

    Styleclipdraw: Coupling content and style in text-to-drawing synthe- sis,

    P . Schaldenbrand, Z. Liu, and J. Oh, “Styleclipdraw: Coupling content and style in text-to-drawing synthe- sis,” arXiv preprint arXiv:2111.03133, 2022. 1, 4

    arXiv 2022

  3. [3]

    Clip-clop: Clip-guided collage and photomontage,

    P . Mirowski, D. Banarse, M. Malinowski, S. Osindero, and C. Fernando, “Clip-clop: Clip-guided collage and photomontage,” arXiv preprint arXiv:2205.03146 , 2022. 1, 4

    arXiv 2022

  4. [4]

    Clipasso: Semantically-aware object sketching,

    Y. Vinker, E. Pajouheshgar, J. Y. Bo, R. C. Bachmann, A. H. Bermano, D. Cohen-Or, A. Zamir, and A. Shamir, “Clipasso: Semantically-aware object sketching,” ACM Transactions on Graphics (TOG), vol. 41, no. 4, pp. 1–11,

    work page

  5. [5]

    Clipascene: Scene sketching with different types and levels of abstraction,

    Y. Vinker, Y. Alaluf, D. Cohen-Or, and A. Shamir, “Clipascene: Scene sketching with different types and levels of abstraction,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) , Oc- tober 2023, pp. 4146–4156. 1

    work page 2023

  6. [6]

    Clipfont: Text guided vector wordart generation,

    Y. Song and Y. Zhang, “Clipfont: Text guided vector wordart generation,” in 33rd British Machine Vision Con- ference 2022, BMVC 2022, London, UK, November 21-24, 2022, 2022. 1

    work page 2022

  7. [7]

    Vectorfusion: Text- to-svg by abstracting pixel-based diffusion models,

    A. Jain, A. Xie, and P . Abbeel, “Vectorfusion: Text- to-svg by abstracting pixel-based diffusion models,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023. 1, 2, 4, 5, 9, 10, 11, 12, 13

    work page 2023

  8. [8]

    Diffsketcher: Text guided vector sketch syn- thesis through latent diffusion models,

    X. Xing, C. Wang, H. Zhou, J. Zhang, Q. Yu, and D. Xu, “Diffsketcher: Text guided vector sketch syn- thesis through latent diffusion models,” in Advances in Neural Information Processing Systems (NeurIPS), 2023. 1, 2, 4, 5, 9, 10, 12, 13

    work page 2023

Show all 68 references
  1. [9]

    Svgdreamer: Text guided svg generation with diffu- sion model,

    X. Xing, H. Zhou, C. Wang, J. Zhang, D. Xu, and Q. Yu, “Svgdreamer: Text guided svg generation with diffu- sion model,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2024, pp. 4546–4555. 1, 2, 10

    2024

  2. [10]

    Supersvg: Superpixel-based scalable vector graphics synthesis,

    T. Hu, R. Yi, B. Qian, J. Zhang, P . L. Rosin, and Y.-K. Lai, “Supersvg: Superpixel-based scalable vector graphics synthesis,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2024, pp. 24 892–24 901. 1, 4

    2024

  3. [11]

    Nivel: Neural implicit vector layers for text-to-vector generation,

    V . Thamizharasan, D. Liu, M. Fisher, N. Zhao, E. Kalogerakis, and M. Lukac, “Nivel: Neural implicit vector layers for text-to-vector generation,” in Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2024, pp. 4589–4597. 1, 4, 13

    2024

  4. [12]

    Text-to-vector gen- eration with neural path representation,

    P . Zhang, N. Zhao, and J. Liao, “Text-to-vector gen- eration with neural path representation,” ACM Trans. Graph., vol. 43, no. 4, Jul. 2024. 1, 4, 13

    2024

  5. [13]

    Differentiable vector graphics rasterization for editing and learning,

    T.-M. Li, M. Luk ´aˇc, G. Micha ¨el, and J. Ragan-Kelley, “Differentiable vector graphics rasterization for editing and learning,” ACM Transactions on Graphics (TOG) , vol. 39, no. 6, pp. 193:1–193:15, 2020. 1, 2, 4

    2020

  6. [14]

    Learning transferable visual models from natu- ral language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, 16 S. Agarwal, G. Sastry, A. Askell, P . Mishkin, J. Clark et al., “Learning transferable visual models from natu- ral language supervision,” in International Conference on Machine Learning (ICML). PMLR, 2021, pp. 8748–8763...

    2021

  7. [15]

    GLIDE: Towards photorealistic image generation and editing with text-guided diffusion models,

    A. Q. Nichol, P . Dhariwal, A. Ramesh, P . Shyam, P . Mishkin, B. Mcgrew, I. Sutskever, and M. Chen, “GLIDE: Towards photorealistic image generation and editing with text-guided diffusion models,” in Pro- ceedings of the 39th International Conference on Machine Learning (ICML)...

    2022

  8. [16]

    High-resolution image synthesis with la- tent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P . Esser, and B. Ommer, “High-resolution image synthesis with la- tent diffusion models,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 10 684–10 695. 1, 4, 6, 9, 11, 14

    2022

  9. [17]

    Hierarchical text-conditional image gener- ation with clip latents,

    A. Ramesh, P . Dhariwal, A. Nichol, C. Chu, and M. Chen, “Hierarchical text-conditional image gener- ation with clip latents,” arXiv preprint arXiv:2204.06125,

    arXiv

  10. [18]

    Photorealistic text- to-image diffusion models with deep language under- standing,

    C. Saharia, W. Chan, S. Saxena, L. Li, J. Whang, E. L. Denton, K. Ghasemipour, R. Gontijo Lopes, B. Karagol Ayan, T. Salimans et al., “Photorealistic text- to-image diffusion models with deep language under- standing,” in Advances in Neural Information Processing Systems (Neur...

    2022

  11. [19]

    If by deepfloyd lab at stabilityai,

    StabilityAI, “If by deepfloyd lab at stabilityai,” https: //github.com/deep-floyd/IF, 2023. 1, 4, 14

    2023

  12. [20]

    Dreamfusion: Text-to-3d using 2d diffusion,

    B. Poole, A. Jain, J. T. Barron, and B. Mildenhall, “Dreamfusion: Text-to-3d using 2d diffusion,” in The Eleventh International Conference on Learning Representa- tions (ICLR), 2023. 1, 2, 4, 5, 12

    2023

  13. [21]

    Word-as-image for semantic typography,

    S. Iluz, Y. Vinker, A. Hertz, D. Berio, D. Cohen-Or, and A. Shamir, “Word-as-image for semantic typography,” ACM Transactions on Graphics (TOG) , vol. 42, no. 4, jul

  14. [22]

    LoRA: Low-rank adaptation of large language models,

    E. J. Hu, yelong shen, P . Wallis, Z. Allen-Zhu, Y. Li, S. Wang, L. Wang, and W. Chen, “LoRA: Low-rank adaptation of large language models,” in International Conference on Learning Representations (ICLR) , 2022. [Online]. Available: https://openreview.net/forum? id=nZeVKeeFYf9 2, 6, 9

    2022

  15. [23]

    Imagereward: Learning and evaluating human preferences for text-to-image generation,

    J. Xu, X. Liu, Y. Wu, Y. Tong, Q. Li, M. Ding, J. Tang, and Y. Dong, “Imagereward: Learning and evaluating human preferences for text-to-image generation,” 2023. 2, 5, 6, 12

    2023

  16. [24]

    Segment anything,

    A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.- Y. Lo et al. , “Segment anything,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2023, pp. 4015–4026. 2, 7

    2023

  17. [25]

    A neural representation of sketch drawings,

    D. Ha and D. Eck, “A neural representation of sketch drawings,” in International Conference on Learning Representations (ICLR), 2018. [Online]. Available: https: //openreview.net/forum?id=Hy6GHpkCW 2

    2018

  18. [26]

    A learned representation for scalable vector graphics,

    R. G. Lopes, D. Ha, D. Eck, and J. Shlens, “A learned representation for scalable vector graphics,” in Proceed- ings of the IEEE/CVF International Conference on Com- puter Vision (ICCV), October 2019. 2

    2019

  19. [27]

    Deepsvg: A hierarchical generative network for vector graphics animation,

    A. Carlier, M. Danelljan, A. Alahi, and R. Timofte, “Deepsvg: A hierarchical generative network for vector graphics animation,” Advances in Neural Information Processing Systems (NeurIPS), vol. 33, pp. 16 351–16 361,

  20. [28]

    Im2vec: Synthesizing vector graphics without vector supervision,

    P . Reddy, M. Gharbi, M. Lukac, and N. J. Mitra, “Im2vec: Synthesizing vector graphics without vector supervision,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp. 7342–7351. 2

    2021

  21. [29]

    Deepvecfont: Synthesizing high- quality vector fonts via dual-modality learning,

    Y. Wang and Z. Lian, “Deepvecfont: Synthesizing high- quality vector fonts via dual-modality learning,” ACM Transactions on Graphics (TOG), vol. 40, no. 6, 2021. 2

    2021

  22. [30]

    Iconshop: Text-based vector icon synthesis with autoregressive transform- ers,

    R. Wu, W. Su, K. Ma, and J. Liao, “Iconshop: Text-based vector icon synthesis with autoregressive transform- ers,” arXiv preprint arXiv:2304.14400, 2023. 2

    2023 arXiv

  23. [31]

    Strokenuwa: Tokenizing strokes for vector graphic synthesis,

    Z. Tang, C. Wu, Z. Zhang, M. Ni, S. Yin, Y. Liu, Z. Yang, L. Wang, Z. Liu, J. Li et al. , “Strokenuwa: Tokenizing strokes for vector graphic synthesis,” arXiv preprint arXiv:2401.17093, 2024. 2

    2024 arXiv

  24. [32]

    Imagenet: A large-scale hierarchical image database,

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, “Imagenet: A large-scale hierarchical image database,” in 2009 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2009, pp. 248–255. 2

    2009

  25. [33]

    Figr: Few-shot image gen- eration with reptile,

    L. Clou ˆatre and M. Demers, “Figr: Few-shot image gen- eration with reptile,” arXiv preprint arXiv:1901.02199 ,

    1901 arXiv

  26. [34]

    Noto emoji fonts,

    Google, “Noto emoji fonts,” https://github.com/ googlefonts/noto-emoji, 2014. 2

    2014

  27. [35]

    Clipgen: A deep generative model for clipart vectorization and synthesis,

    I.-C. Shen and B.-Y. Chen, “Clipgen: A deep generative model for clipart vectorization and synthesis,” IEEE Transactions on Visualization and Computer Graphics, vol. 28, no. 12, p. 4211–4224, dec 2022. [Online]. Available: https://doi.org/10.1109/TVCG. 2021.3084944 4

    2022

  28. [36]

    Modern evolution strategies for cre- ativity: Fitting concrete images and abstract concepts,

    Y. Tian and D. Ha, “Modern evolution strategies for cre- ativity: Fitting concrete images and abstract concepts,” in Artificial Intelligence in Music, Sound, Art and Design . Springer, 2022, pp. 275–291. 4, 9, 10

    2022

  29. [37]

    Towards layer-wise image vectorization,

    X. Ma, Y. Zhou, X. Xu, B. Sun, V . Filev, N. Orlov, Y. Fu, and H. Shi, “Towards layer-wise image vectorization,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 16 314– 16 323. 4, 5, 11

    2022

  30. [38]

    Marvel: Raster gray-level manga vectoriza- tion via primitive-wise deep reinforcement learning,

    H. Su, X. Liu, J. Niu, J. Cui, J. Wan, X. Wu, and N. Wang, “Marvel: Raster gray-level manga vectoriza- tion via primitive-wise deep reinforcement learning,” IEEE Transactions on Circuits and Systems for Video Tech- nology (T-CSVT), 2023. 4

    2023

  31. [39]

    Clipvg: Text-guided image manipulation using differentiable vector graphics,

    Y. Song, X. Shao, K. Chen, W. Zhang, Z. Jing, and M. Li, “Clipvg: Text-guided image manipulation using differentiable vector graphics,” in Proceedings of the Conference on Artificial Intelligence (AAAI), 2023. 4

    2023

  32. [40]

    Deep unsupervised learning using nonequilibrium thermodynamics,

    J. Sohl-Dickstein, E. Weiss, N. Maheswaranathan, and S. Ganguli, “Deep unsupervised learning using nonequilibrium thermodynamics,” in Proceedings of the International Conference on Machine Learning (ICML) , vol. 37, 2015, pp. 2256–2265. 4

    2015

  33. [41]

    Generative modeling by esti- mating gradients of the data distribution,

    Y. Song and S. Ermon, “Generative modeling by esti- mating gradients of the data distribution,” in Advances in Neural Information Processing Systems (NeurIPS) , 17 vol. 32, 2019. 4

    2019

  34. [42]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P . Abbeel, “Denoising diffusion probabilistic models,” in Advances in Neural Information Processing Systems (NeurIPS) , vol. 33, 2020, pp. 6840–

    2020

  35. [43]

    Score-based generative mod- eling through stochastic differential equations,

    Y. Song, J. Sohl-Dickstein, D. P . Kingma, A. Kumar, S. Ermon, and B. Poole, “Score-based generative mod- eling through stochastic differential equations,” in In- ternational Conference on Learning Representations (ICLR),

  36. [44]

    Diffusion models beat gans on image synthesis,

    P . Dhariwal and A. Nichol, “Diffusion models beat gans on image synthesis,”Advances in neural information processing systems (NeurIPS) , vol. 34, pp. 8780–8794,

  37. [45]

    Improved denoising diffusion probabilistic models,

    A. Q. Nichol and P . Dhariwal, “Improved denoising diffusion probabilistic models,” in International confer- ence on machine learning (ICLR), 2021, pp. 8162–8171. 4

    2021

  38. [46]

    Denoising diffusion implicit models,

    J. Song, C. Meng, and S. Ermon, “Denoising diffusion implicit models,” in International Conference on Learning Representations (ICLR), 2021. 4, 6

    2021

  39. [47]

    Classifier-free diffusion guid- ance,

    J. Ho and T. Salimans, “Classifier-free diffusion guid- ance,” arXiv preprint arXiv:2207.12598, 2022. 4, 10

    2022 arXiv

  40. [48]

    SDXL: Improving latent diffusion models for high-resolution image synthesis,

    D. Podell, Z. English, K. Lacey, A. Blattmann, T. Dock- horn, J. M ¨uller, J. Penna, and R. Rombach, “SDXL: Improving latent diffusion models for high-resolution image synthesis,” in The Twelfth International Conference on Learning Representations (ICLR), 2024. 4, 14

    2024

  41. [49]

    Score jacobian chaining: Lifting pretrained 2d diffu- sion models for 3d generation,

    H. Wang, X. Du, J. Li, R. A. Yeh, and G. Shakhnarovich, “Score jacobian chaining: Lifting pretrained 2d diffu- sion models for 3d generation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2023, pp. 12 619–12 629. 4

    2023

  42. [50]

    Video diffusion models,

    J. Ho, T. Salimans, A. Gritsenko, W. Chan, M. Norouzi, and D. J. Fleet, “Video diffusion models,” Advances in Neural Information Processing Systems (NeurIPS), vol. 35, pp. 8633–8646, 2022. 4

    2022

  43. [51]

    Make-a-video: Text-to- video generation without text-video data,

    U. Singer, A. Polyak, T. Hayes, X. Yin, J. An, S. Zhang, Q. Hu, H. Yang, O. Ashual, O. Gafni, D. Parikh, S. Gupta, and Y. Taigman, “Make-a-video: Text-to- video generation without text-video data,” in The Eleventh International Conference on Learning Representa- tions (ICLR), 2023. 4

    2023

  44. [52]

    Clip-forge: Towards zero-shot text-to-shape generation,

    A. Sanghi, H. Chu, J. G. Lambourne, Y. Wang, C.-Y. Cheng, M. Fumero, and K. R. Malekshan, “Clip-forge: Towards zero-shot text-to-shape generation,” in Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 18 603–18 613. 4

    2022

  45. [53]

    Zero-shot text-guided object generation with dream fields,

    A. Jain, B. Mildenhall, J. T. Barron, P . Abbeel, and B. Poole, “Zero-shot text-guided object generation with dream fields,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (CVPR) , 2022, pp. 867–876. 4

    2022

  46. [54]

    Do 2d {gan}s know 3d shape? unsupervised 3d shape reconstruction from 2d image {gan}s,

    X. Pan, B. Dai, Z. Liu, C. C. Loy, and P . Luo, “Do 2d {gan}s know 3d shape? unsupervised 3d shape reconstruction from 2d image {gan}s,” in International Conference on Learning Representations (ICLR) , 2021. [Online]. Available: https://openreview.net/forum? id=FGqiDsBUKL0 4

    2021

  47. [55]

    Magic3d: High-resolution text-to-3d content cre- ation,

    C.-H. Lin, J. Gao, L. Tang, T. Takikawa, X. Zeng, X. Huang, K. Kreis, S. Fidler, M.-Y. Liu, and T.-Y. Lin, “Magic3d: High-resolution text-to-3d content cre- ation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , June 2023, pp. 300–309. 4

    2023

  48. [56]

    Prolificdreamer: High-fidelity and diverse text-to-3d generation with variational score distillation,

    Z. Wang, C. Lu, Y. Wang, F. Bao, C. Li, H. Su, and J. Zhu, “Prolificdreamer: High-fidelity and diverse text-to-3d generation with variational score distillation,” arXiv preprint arXiv:2305.16213, 2023. 4, 5, 6, 12

    2023 arXiv

  49. [57]

    Fantasia3d: Disentangling geometry and appearance for high- quality text-to-3d content creation,

    R. Chen, Y. Chen, N. Jiao, and K. Jia, “Fantasia3d: Disentangling geometry and appearance for high- quality text-to-3d content creation,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), October 2023. 4

    2023

  50. [58]

    Taming trans- formers for high-resolution image synthesis,

    P . Esser, R. Rombach, and B. Ommer, “Taming trans- formers for high-resolution image synthesis,” in Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition (CVPR) , 2021, pp. 12 873–12 883. 5

    2021

  51. [59]

    Grounded sam: Assembling open-world models for diverse visual tasks,

    T. Ren, S. Liu, A. Zeng, J. Lin, K. Li, H. Cao, J. Chen, X. Huang, Y. Chen, F. Yan, Z. Zeng, H. Zhang, F. Li, J. Yang, H. Li, Q. Jiang, and L. Zhang, “Grounded sam: Assembling open-world models for diverse visual tasks,” 2024. 7

    2024

  52. [60]

    Gans trained by a two time-scale update rule converge to a local nash equilibrium,

    M. Heusel, H. Ramsauer, T. Unterthiner, B. Nessler, and S. Hochreiter, “Gans trained by a two time-scale update rule converge to a local nash equilibrium,” Advances in neural information processing systems (NeurIPS) , vol. 30,

  53. [61]

    Image quality metrics: Psnr vs. ssim,

    A. Hor ´e and D. Ziou, “Image quality metrics: Psnr vs. ssim,” in 2010 20th International Conference on Pattern Recognition, 2010, pp. 2366–2369. 9

    2010

  54. [62]

    Blip: Bootstrap- ping language-image pre-training for unified vision- language understanding and generation,

    J. Li, D. Li, C. Xiong, and S. Hoi, “Blip: Bootstrap- ping language-image pre-training for unified vision- language understanding and generation,” in Interna- tional Conference on Machine Learning (ICML) . PMLR, 2022, pp. 12 888–12 900. 9

    2022

  55. [63]

    Improved aesthetic predictor,

    C. Schuhmann, “Improved aesthetic predictor,” https://github.com/christophschuhmann/ improved-aesthetic-predictor, 2022. 9

    2022

  56. [64]

    Human preference score: Better aligning text-to-image models with human preference,

    X. Wu, K. Sun, F. Zhu, R. Zhao, and H. Li, “Human preference score: Better aligning text-to-image models with human preference,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) , Oc- tober 2023, pp. 2096–2105. 9

    2023

  57. [65]

    Resolution-robust large mask in- painting with fourier convolutions,

    R. Suvorov, E. Logacheva, A. Mashikhin, A. Remizova, A. Ashukha, A. Silvestrov, N. Kong, H. Goka, K. Park, and V . Lempitsky, “Resolution-robust large mask in- painting with fourier convolutions,” arXiv preprint arXiv:2109.07161, 2021. 9

    2021 arXiv

  58. [66]

    Glyphcontrol: glyph conditional control for visual text generation,

    Y. Yang, D. Gui, Y. Yuan, W. Liang, H. Ding, H. Hu, and K. Chen, “Glyphcontrol: glyph conditional control for visual text generation,” in Proceedings of the 37th International Conference on Neural Information Processing Systems (NIPS), 2024. 14

    2024

  59. [67]

    Textdiffuser: Diffusion models as text painters,

    J. Chen, Y. Huang, T. Lv, L. Cui, Q. Chen, and F. Wei, “Textdiffuser: Diffusion models as text painters,” arXiv preprint arXiv:2305.10855, 2023. 14

    2023 arXiv

  60. [68]

    Glyph-byt5-v2: A strong aesthetic baseline for accu- rate multilingual visual text rendering,

    Z. Liu, W. Liang, Y. Zhao, B. Chen, J. Li, and Y. Yuan, “Glyph-byt5-v2: A strong aesthetic baseline for accu- rate multilingual visual text rendering,” arXiv preprint arXiv:2406.10208, 2024. 14

    2024 arXiv

Pith tools

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