Pith. sign in

REVIEW 4 major objections 6 minor 53 references

Emergence of Painting Ability via Recognition-Driven Evolution

T0 review · 4 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read A vector painting optimized solely for recognition accuracy develops human-like painting ability, beating CLIPasso and compressing images to 0.059 bits per pixel.

desk verdict A legitimate CLIPasso-plus-color-palette extension whose headline 'emergence' and recognition-superiority claims are undercut by evaluating with the same CLIP model used as the optimization objective. read the letter →

arxiv 2501.04966 v1 pith:SGPUOROF submitted 2025-01-09 cs.CV

Yi Lin , Lin Gu , Ziteng Cui , Shenghan Su , Yumo Hao , Yingtao Tian , Tatsuya Harada , Jianfei Yang This is my paper · ORCID
classification cs.CV
keywords paintinggenerationvectorsketchBézierstrokesCLIPrecognition-drivenoptimizationvisualcommunicationefficiencyimagecompressionscenariocomplexity
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 tries to establish that a machine can develop human-like painting ability without imitating human art, by treating recognition accuracy as the sole evolutionary pressure. Its model represents paintings as Bézier-curve strokes, splits them into black and colored sets, learns a restricted palette, and optimizes stroke parameters so that a pre-trained CLIP image encoder gives the painting an embedding close to the original image. The authors report that this procedure yields higher CLIP zero-shot recognition than the CLIPasso baseline at 32 strokes (top-3 accuracy 50.53% versus 48.42%), with recognition accuracy rising as stroke count grows. They also report that the vector paintings compress images to 0.059 bits per pixel while retaining recognition accuracy near JPEG at 0.433 bits per pixel. A sympathetic reader would care because the claim is that minimal, parameter-free artistic skill can emerge from an efficiency objective rather than from style supervision.

What carries the argument

The load-bearing object is the stroke-and-palette painting model built on a differentiable rasterizer (DiffVG), which lets gradients flow from the recognition loss back through rasterization into Bézier control points, widths, and RGB colors. The Stroke Branch optimizes one black set and one colored set of strokes simultaneously, and the Palette Branch learns a palette $P(I)\in\mathbb{R}^{C\times3}$ by localizing points in RGB space through cross-attention queries. The recognition module supplies the evolutionary pressure: $\mathcal{L}_{\text{semantic}} = \sum_l \cos(\phi(I), \phi(S))$ pushes the generated painting's CLIP embedding toward the original image's embedding, $\mathcal{L}_{\text{structure}}$ preserves low-level shape, and $\mathcal{L}_{\text{colour}}$ fits the reference palette. This joint loss is the mechanism that translates 'communicate with minimal strokes and colors' into concrete stroke adjustments.

What would settle it

Evaluate the generated paintings with a recognition encoder that was not used in the semantic loss—for example, a separately trained vision transformer or a human-labeled sketch recognition benchmark—and check whether accuracy against that held-out evaluator still exceeds the CLIPasso baseline.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the combination of a differentiable stroke renderer, a learned color palette, and a recognition-driven loss is sufficient for human-like painting ability to emerge. The model renders each painting as two overlays: a black sketch and a colored sketch, both composed of Bézier strokes. A palette branch predicts C colors from the image through a transformer cross-attention module, and the whole system is optimized with a semantic loss $\sum_l \cos(\phi(I), \phi(S))$ plus structural and color losses. The paper reports that the final painting S reaches 50.53% top-3 zero-shot recognition accuracy at 32 strokes, surpassing CLIPasso's 48.42%, while also achieving aesthetic scores above those of either stroke set alone. The same representation yields a low-bit-rate compression result of 0.059 bpp with 74.51% recognition accuracy against JPEG's 0.433 bpp and 79.10%.

Load-bearing premise

The load-bearing premise is that a CLIP embedding's cosine similarity is the right measure of visual communication efficiency, and that recognition accuracy measured with that same CLIP model is an independent outcome; if that similarity is only a proxy, the 'emergence' is really an optimization artifact.

Editorial extensions

If this is right

  • Increasing the stroke budget raises recognition accuracy monotonically in the reported range, from 7.89% top-3 at 4 strokes to 50.53% at 32 strokes for the combined painting S.
  • Combining black and colored strokes outperforms each alone: S reaches 50.53% top-3 at 32 strokes, while Sblack and Scolour reach 48.42% and 39.47% respectively.
  • The model's aesthetic score improves with stroke count, and the combined painting scores above its black-only and color-only variants on COCO and PASCAL VOC images.
  • The vector representation functions as a low-bit-rate compression: 64 total strokes at 0.059 bpp yields 74.51% recognition accuracy, close to JPEG's 79.10% at 0.433 bpp.
  • A scenario complexity estimator built on CLIP and the Sketchy dataset can select images that resemble historical painting scenes across complexity levels.

Reading between the lines

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

  • The 'evolutionary' framing is an analogy: the actual mechanism is gradient descent on a fixed objective, not population-level variation and selection; a stronger test would let stroke counts and palette sizes vary dynamically under a fitness function.
  • Because recognition is measured with the same CLIP encoder used in the loss, the headline numbers likely overstate general painting ability; a natural extension is to hold out the evaluator, or to test human recognition of the paintings.
  • The compression comparison is task-specific: 0.059 bpp with recognition accuracy is not the same as reconstruction fidelity; pairing this vector representation with a learned decoder could make the compression claim testable on standard image-quality metrics.
  • The reported stylistic resemblance to medieval art is post-hoc interpretation by the authors, not a measured property; a reader could test it with a controlled human study asking artists to date or attribute the generated paintings.
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.

Referee Report

4 major / 6 minor

Summary. This paper proposes an optimization-based painting model that renders an input image as a set of Bézier-curve strokes, split into black and colored stroke groups, with a palette branch that learns a limited color palette. The loss combines a CLIP-based semantic term, a CLIP layer-wise structural term, and a color-palette term. The authors also introduce a scenario-complexity estimation system and a dataset of about 500 images selected from ImageNet, COCO, and PASCAL VOC to mimic artistic scenarios. Experiments report zero-shot CLIP recognition accuracy of generated paintings at 4, 8, 16, and 32 strokes, aesthetic scores, comparisons with CLIPasso, and a low-bit-rate compression demonstration. The paper argues that painting ability emerges from evolutionary pressure toward visual communication efficiency.

Significance. The method is technically concrete and the paper has useful ingredients: a differentiable rasterization pipeline with separate black and colored stroke branches, a compact palette-learning mechanism, a new complexity-filtered painting dataset, and comparisons with CLIPasso. The artistic analysis in Sec. 6 is thoughtful and could be of interest to the sketch and painting community. However, the central quantitative claim—that recognition accuracy improves because of emergent communication efficiency—is weakened by the fact that the same CLIP encoder is used both as the optimization objective and as the evaluation metric, and by the small, unpaired numerical differences. The method is best viewed, with the current evidence, as a CLIP-space vector-painting optimizer; independent evaluation would be needed to substantiate the emergence framing. Because the required fixes are substantial but feasible, the paper warrants major revision.

major comments (4)
  1. [Sec. 3.4 and Sec. 5.1] The recognition accuracy reported in Table 1 is not an independent outcome: L_semantic in Sec. 3.4 minimizes cosine(phi(I), phi(S)) using a pre-trained CLIP image encoder, and the zero-shot evaluation in Sec. 5.1 computes classification with the same CLIP ViT-B/32 encoder and prompts projected into the same space. Table 1 therefore largely verifies that the optimization increased CLIP cosine similarity to the source image, rather than demonstrating that the paintings are generally recognizable. I request evaluation with a different encoder (for example, CLIP ViT-L/14 or SigLIP), or with a classifier or human annotator not used in the loss, plus a direct report of CLIP cosine similarity to the source image as a sanity check; the emergence language should be softened unless such independent evidence is provided.
  2. [Table 1 (Sec. 5.1)] The headline advantage of S over CLIPasso at 32 strokes is 50.53% versus 48.42% top-3 accuracy, which is about 4 images out of 190. The manuscript reports no error bars, no multiple seeds, no statistical test, and no paired analysis; the Sblack row ties CLIPasso exactly at 48.42%, so the entire claimed benefit rests on the colored-stroke addition on a handful of images. Please add confidence intervals, repeated runs, and a paired significance test (for example, McNemar) before claiming superior performance in high-level recognition tasks.
  3. [Sec. 5.2, Fig. 7] The image-compression claim is supported by a single illustrative example: 0.059 bpp versus JPEG at 0.433 bpp and accuracies 74.51% versus 79.10% for one image. This does not substantiate the claim of outperforming traditional methods. A proper evaluation needs a rate-distortion or accuracy-rate comparison over the full evaluation set, with multiple compression ratios, standard codecs such as JPEG, WebP, and BPG, and a clear definition of how the vector painting is converted to bits, including control-point coordinates, colors, widths, and palette storage.
  4. [Table 2] In Table 2, two rows are both labeled Sblack but contain different numeric values; presumably one of them should be Scolour. This obscures the aesthetic-score comparison that supports the claim that colored strokes improve aesthetics. Please correct the labels and report scores for S, Sblack, and Scolour separately, ideally with standard deviations across images.
minor comments (6)
  1. [Sec. 3.4] The equation for L_semantic is written with a sum over layers but without layer indices in the cosine term, while L_structure explicitly uses phi_l; this should be made consistent and the final loss equation should be numbered for reference.
  2. [Sec. 5.1] The text says that optimization leads to higher accuracy in the final painting Scolour, but Table 1 reports the combined painting as row S; please clarify the naming of S, Sblack, and Scolour throughout the section.
  3. [References] Reference [34] is a placeholder (Author Name, Project title, Year); a full citation for the aesthetic indicator is needed.
  4. [Sec. 5.1] There is a broken figure reference (Figure ??) and a missing reference to Figure 5 in the text; these should be corrected.
  5. [General] There are several typos and grammatical errors, for example 'approch' in Sec. 2.1, 'amgutity' in Sec. 4.1, and 'evaluate' in Sec. 5.1; a careful proofreading pass is needed.
  6. [Sec. 5.2] The text says more comparison results with traditional compression methods are shown in the discussion section, but those results do not appear in the main text; either add them or remove the pointer.

Circularity Check

1 steps flagged · score 7.0 of 10

Reported recognition gains are measured with the same CLIP encoder used as the optimization objective, so the headline 'emergent' recognition claim is not independently established.

  1. fitted input called prediction [Sec. 3.4 (Recognition Module, L_semantic) and Sec. 5.1 (Quantitative Evaluation, Table 1)]
    "The semantic loss Lsemantic is central to our method, as it reflects the recognition accuracy of the generated painting. It is computed by measuring the cosine similarity between the high-level embeddings of the generated painting and the input image. ... Specifically, recognition accuracy indicates how well a painting can be identified by a machine as belonging to the same category as the original image."

    L_semantic is minimized by moving the CLIP embedding phi(S) of the painting toward the CLIP embedding phi(I) of the input image. The zero-shot evaluation then classifies the painting using the same pre-trained CLIP encoder, scoring cos(phi(S), text(class)). When the original image is correctly classified by CLIP, cos(phi(I), correct_text) exceeds the other text scores, and driving phi(S) toward phi(I) transfers that margin to the painting by construction. Table 1 therefore reports how well the optimizer satisfied its own CLIP-cosine objective, not recognition by an independent system; it is like reporting training accuracy as evidence of an emergent capability.

full rationale

The central quantitative evidence for 'emergence' is a closed loop: the strokes are optimized against CLIP cosine similarity to the source image, and the recognition metric is CLIP zero-shot accuracy on the same encoder. No independent recognizer, held-out evaluator, or non-CLIP measure is used to confirm that the paintings are recognizable as their categories. The comparison to CLIPasso is a comparison of two CLIP-space optimizers and does not establish emergent communicative value. The paper also reports aesthetic scores from a separate CLIP-based aesthetic predictor and a subjective art-historical analysis; those are not circular in the same way, but they do not rescue the recognition claim. The author self-citations (e.g., [8], [40], [49]) are used for technique or motivation and are not load-bearing. The 50.53% vs 48.42% top-3 difference at 32 strokes is also only about 4 images out of 190 with no error bars, which makes the headline margin fragile, but the main issue is the evaluator/objective identity. The method may be a legitimate CLIP-space painting and compression optimizer, but the 'emergent human-like painting ability' claim is not independently supported.

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

The paper does not introduce new physical entities. The 'Reference Palette Queries' are learnable parameters, not a new entity. The main hidden assumptions are the validity of CLIP as a proxy for visual communication and the transferability of the Sketchy complexity estimator.

free parameters (5)
  • Loss weights lambda1, lambda2, lambda3 = 1.0, 1.0, 1.0
    Set to 1.0 without ablation in Sec. 3.4; the balance between structural, semantic, and color losses is not explored.
  • Number of strokes N = 4, 8, 16, 32
    Chosen by hand for experiments; the model does not automatically determine the stroke count.
  • Number of colors C = up to 16 (4 or 8 for low stroke counts)
    The palette size is manually specified per experiment in Sec. 5.1.
  • Stroke width and opacity = width 1.0 for black, opacity 0.5 for colored
    Fixed constants in Sec. 3.2, not learned or ablated.
  • Top-10 threshold in Composition Criterion = 10
    Used to select images similar to reference paintings in Sec. 4.2; no sensitivity analysis.
assumptions (4)
  • domain assumption CLIP embeddings capture semantic content sufficient for recognition and painting optimization.
    The entire method relies on CLIP as the measure of semantic similarity and as the recognition module in Sec. 3.4 and Sec. 5.1.
  • standard math The differentiable rasterizer diffVG provides accurate gradients for stroke optimization.
    The method builds on diffVG [30] without questioning its accuracy or limitations.
  • domain assumption The Sketchy dataset complexity labels transfer to natural-image painting scenarios.
    The scenario complexity estimator is trained on Sketchy [37] and then applied to ImageNet, COCO, and VOC images in Sec. 4.3; this cross-domain transfer is assumed valid.
  • domain assumption A painting that maximizes CLIP similarity to the input image also achieves high zero-shot recognition accuracy.
    This is the core premise of the 'visual communication efficiency' framing, but it is an assumption that is not independently verified.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Emergence of Painting Ability via Recognition-Driven Evolution." pith.science (2026). https://pith.science/paper/SGPUOROF

@misc{pith2026250104966,
  author       = {Pith},
  title        = {Pith review of: Emergence of Painting Ability via Recognition-Driven Evolution},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SGPUOROF}},
  note         = {Machine review of arXiv:2501.04966}
}
read the original abstract

From Paleolithic cave paintings to Impressionism, human painting has evolved to depict increasingly complex and detailed scenes, conveying more nuanced messages. This paper attempts to emerge this artistic capability by simulating the evolutionary pressures that enhance visual communication efficiency. Specifically, we present a model with a stroke branch and a palette branch that together simulate human-like painting. The palette branch learns a limited colour palette, while the stroke branch parameterises each stroke using B\'ezier curves to render an image, subsequently evaluated by a high-level recognition module. We quantify the efficiency of visual communication by measuring the recognition accuracy achieved with machine vision. The model then optimises the control points and colour choices for each stroke to maximise recognition accuracy with minimal strokes and colours. Experimental results show that our model achieves superior performance in high-level recognition tasks, delivering artistic expression and aesthetic appeal, especially in abstract sketches. Additionally, our approach shows promise as an efficient bit-level image compression technique, outperforming traditional methods.

Figures

Figures reproduced from arXiv: 2501.04966 by the authors.

Figure 1
Figure 1. From cave paintings to Impressionism: the evolution of [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of the painting model. The model generates human-like paintings using a Stroke Branch for stroke optimization [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. The process of stroke initialization and the generation of [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (14 more)
Figure 4
Figure 4. Figure 4: Overview of the scenario complexity estimation system. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Generated paintings with different numbers of strokes. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Top-3 predictions and the corresponding recognition accuracy of the generated paintings. Right: examples which demonstrate [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Comparison of image quality and recognition accuracy [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Line drawing in ancient China. There is a slight transi [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 9
Figure 9. Figure 9: Illustration of Medieval Painting Art.(a)-(b) Illustrations [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]
Figure 10
Figure 10. Figure 10: Paintings of Bird. These generated paintings demonstrate how the line drawing technique unique to ancient Chinese art is [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]
Figure 11
Figure 11. Figure 11: Paintings of Church and Stupa. These generated paintings demonstrate how the flattened and two-dimensional drawing style [PITH_FULL_IMAGE:figures/full_fig_p010_11.png]
Figure 12
Figure 12. Figure 12: More painting of architecture. Examples are selected from the ”triumphal arch” and ”beacon” categories. [PITH_FULL_IMAGE:figures/full_fig_p011_12.png]
Figure 13
Figure 13. Figure 13: Paintings of Birds, Elephants, and Cows. [PITH_FULL_IMAGE:figures/full_fig_p012_13.png]
Figure 14
Figure 14. Figure 14: Depiction of birds in Western Art. (a) Birds in Lascaux [PITH_FULL_IMAGE:figures/full_fig_p012_14.png]
Figure 17
Figure 17. Figure 17: Depiction of cows in Cave Paintings. (a) [PITH_FULL_IMAGE:figures/full_fig_p013_17.png]
Figure 15
Figure 15. Figure 15: Mogao Cave Paintings. (a) Waterfowl, Cave 285, West￾ern Wei Dynasty. (b) White Goose in Lotus Pond, Cave 435, Northern Wei Dynasty. (c) Baboon and Gentoo, Cave 249, West￾ern Wei Dynasty. (a) shows a line drawing of the bird before col￾oring, while (b)-(c) demonstrate …
Figure 16
Figure 16. Figure 16: More Mogao Cave Paintings. (a) The process of Dun [PITH_FULL_IMAGE:figures/full_fig_p013_16.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 36 canonical work pages

  1. [1]

    Generative adversar- ial networks for extreme learned image compression

    Eirikur Agustsson, Michael Tschannen, Fabian Mentzer, Radu Timofte, and Luc Van Gool. Generative adversar- ial networks for extreme learned image compression. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 221–231, 2019. 3

    work page 2019

  2. [2]

    Multi-realism image compression with a conditional generator

    Eirikur Agustsson, David Minnen, George Toderici, and Fabian Mentzer. Multi-realism image compression with a conditional generator. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 22324–22333, 2023. 4

    work page 2023

  3. [3]

    An image is worth 16x16 words: Trans- formers for image recognition at scale

    Dosovitskiy Alexey. An image is worth 16x16 words: Trans- formers for image recognition at scale. arXiv preprint arXiv: 2010.11929, 2020. 4

    arXiv 2010

  4. [4]

    Basic color terms: Their univer- sality and evolution

    Brent Berlin and Paul Kay. Basic color terms: Their univer- sality and evolution. Univ of California Press, 1991. 2

    work page 1991

  5. [5]

    End-to- end object detection with transformers

    Nicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, and Sergey Zagoruyko. End-to- end object detection with transformers. In European confer- ence on computer vision, pages 213–229. Springer, 2020. 5

    2020

  6. [6]

    Communicating artificial neural net- works develop efficient color-naming systems

    Rahma Chaabouni, Eugene Kharitonov, Emmanuel Dupoux, and Marco Baroni. Communicating artificial neural net- works develop efficient color-naming systems. Proceedings of the National Academy of Sciences, 118(12):e2016569118,

    work page

  7. [7]

    Learning to generate line drawings that convey geometry and semantics

    Caroline Chan, Fr ´edo Durand, and Phillip Isola. Learning to generate line drawings that convey geometry and semantics. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 7915–7925, 2022. 3

    work page 2022

  8. [8]

    You only need 90k parameters to adapt light: a light weight trans- former for image enhancement and exposure correction

    Ziteng Cui, Kunchang Li, Lin Gu, Shenghan Su, Peng Gao, Zhengkai Jiang, Yu Qiao, and Tatsuya Harada. You only need 90k parameters to adapt light: a light weight trans- former for image enhancement and exposure correction. arXiv preprint arXiv:2205.14871, 2022. 5

    arXiv 2022

Show all 53 references
  1. [9]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, pages 248–255. Ieee, 2009. 6

    2009

  2. [10]

    Edge augmentation for large-scale sketch recognition with- out sketches

    Nikos Efthymiadis, Giorgos Tolias, and Ond ˇrej Chum. Edge augmentation for large-scale sketch recognition with- out sketches. In 2022 26th International Conference on Pat- tern Recognition (ICPR), pages 3595–3602. IEEE, 2022. 3

    2022

  3. [11]

    How do humans sketch objects? ACM Trans

    Mathias Eitz, James Hays, and Marc Alexa. How do humans sketch objects? ACM Trans. Graph. (Proc. SIGGRAPH), 31 (4):44:1–44:10, 2012. 3

    2012

  4. [12]

    The pascal visual object classes (voc) challenge

    Mark Everingham, Luc Van Gool, Christopher KI Williams, John Winn, and Andrew Zisserman. The pascal visual object classes (voc) challenge. International journal of computer vision, 88:303–338, 2010. 6

    2010

  5. [13]

    An adaptive algorithm for spatial gray- scale

    Robert W Floyd. An adaptive algorithm for spatial gray- scale. In Proc. Soc. Inf. Disp., pages 75–77, 1976. 8

    1976

  6. [14]

    Clipdraw: Exploring text-to-drawing synthesis through language-image encoders

    Kevin Frans, Lisa Soros, and Olaf Witkowski. Clipdraw: Exploring text-to-drawing synthesis through language-image encoders. Advances in Neural Information Processing Sys- tems, 35:5207–5218, 2022. 2, 3, 5

    2022

  7. [15]

    Sketchycoco: Image gener- ation from freehand scene sketches

    Chengying Gao, Qi Liu, Qi Xu, Limin Wang, Jianzhuang Liu, and Changqing Zou. Sketchycoco: Image gener- ation from freehand scene sketches. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5174–5183, 2020. 3

    2020

  8. [16]

    A simple method for color quantization: Octree quantization

    Michael Gervautz and Werner Purgathofer. A simple method for color quantization: Octree quantization. In New Trends in Computer Graphics: Proceedings of CG International’88, pages 219–231. Springer, 1988. 8

    1988

  9. [17]

    Color naming across languages reflects color use

    Edward Gibson, Richard Futrell, Julian Jara-Ettinger, Kyle Mahowald, Leon Bergen, Sivalogeswaran Ratnasingam, Mitchell Gibson, Steven T Piantadosi, and Bevil R Conway. Color naming across languages reflects color use. Proceed- ings of the National Academy of Sciences , 114(40)...

    2017

  10. [18]

    Primal sketch: Integrating structure and texture

    Cheng-en Guo, Song-Chun Zhu, and Ying Nian Wu. Primal sketch: Integrating structure and texture. Computer Vision and Image Understanding, 106(1):5–19, 2007. 4

    2007

  11. [19]

    Color image quantization for frame buffer display

    Paul Heckbert. Color image quantization for frame buffer display. ACM Siggraph Computer Graphics, 16(3):297–307,

  12. [20]

    Gaussian error linear units (gelus)

    Dan Hendrycks and Kevin Gimpel. Gaussian error linear units (gelus). arXiv preprint arXiv:1606.08415, 2016. 5

    2016 arXiv

  13. [21]

    Vectorpainter: A novel approach to stylized vector graphics synthesis with vectorized strokes

    Juncheng Hu, Ximing Xing, Zhengqi Zhang, Jing Zhang, and Qian Yu. Vectorpainter: A novel approach to stylized vector graphics synthesis with vectorized strokes. arXiv preprint arXiv:2405.02962, 2024. 3

    2024 arXiv

  14. [22]

    Image-to-image translation with conditional adver- sarial networks

    Phillip Isola, Jun-Yan Zhu, Tinghui Zhou, and Alexei A Efros. Image-to-image translation with conditional adver- sarial networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1125–1134,

  15. [23]

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

    Ajay Jain, Amber Xie, and Pieter Abbeel. Vectorfusion: Text-to-svg by abstracting pixel-based diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 1911–1920, 2023. 3

    1911

  16. [24]

    Synthesizing human- like sketches from natural images using a conditional con- volutional decoder

    Moritz Kampelmuhler and Axel Pinz. Synthesizing human- like sketches from natural images using a conditional con- volutional decoder. In Proceedings of the IEEE/CVF winter conference on applications of computer vision, pages 3203– 3211, 2020. 2, 3

    2020

  17. [25]

    Diffu- sionclip: Text-guided diffusion models for robust image ma- nipulation

    Gwanghyun Kim, Taesung Kwon, and Jong Chul Ye. Diffu- sionclip: Text-guided diffusion models for robust image ma- nipulation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2426–2435,

  18. [26]

    Egic: Enhanced low-bit-rate generative image compression guided by semantic segmentation

    Nikolai K ¨orber, Eduard Kromer, Andreas Siebert, Sascha Hauke, Daniel Mueller-Gritschneder, and Bj ¨orn Schuller. Egic: Enhanced low-bit-rate generative image compression guided by semantic segmentation. In European Conference on Computer Vision, pages 202–220. Springer, 2025. 4

    2025

  19. [27]

    Learning geometry- aware representations by sketching

    Hyundo Lee, Inwoo Hwang, Hyunsung Go, Won-Seok Choi, Kibeom Kim, and Byoung-Tak Zhang. Learning geometry- aware representations by sketching. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 23315–23326, 2023. 3

    2023

  20. [28]

    Pose recognition with cascade transformers

    Ke Li, Shijie Wang, Xiang Zhang, Yifan Xu, Weijian Xu, and Zhuowen Tu. Pose recognition with cascade transformers. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 1944–1953, 2021. 5

    1944

  21. [29]

    Photo-sketching: Inferring contour draw- ings from images

    Mengtian Li, Zhe Lin, Radomir Mech, Ersin Yumer, and Deva Ramanan. Photo-sketching: Inferring contour draw- ings from images. In 2019 IEEE Winter Conference on Ap- plications of Computer Vision (WACV) , pages 1403–1412. IEEE, 2019. 2, 3

    2019

  22. [30]

    Differentiable vector graphics rasterization for editing and learning

    Tzu-Mao Li, Michal Luk ´aˇc, Micha ¨el Gharbi, and Jonathan Ragan-Kelley. Differentiable vector graphics rasterization for editing and learning. ACM Transactions on Graphics (TOG), 39(6):1–15, 2020. 3, 4

    2020

  23. [31]

    Im2pencil: Controllable pencil illustra- tion from photographs

    Yijun Li, Chen Fang, Aaron Hertzmann, Eli Shechtman, and Ming-Hsuan Yang. Im2pencil: Controllable pencil illustra- tion from photographs. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 1525–1534, 2019. 3

    2019

  24. [32]

    Microsoft coco: Common objects in context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll´ar, and C Lawrence Zitnick. Microsoft coco: Common objects in context. In Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceeding...

    2014

  25. [33]

    A computational investigation into the human representation and processing of visual information

    D Man and A Vision. A computational investigation into the human representation and processing of visual information. WH San Francisco: Freeman and Company, San Francisco, 1:1, 1982. 4

    1982

  26. [34]

    Project title, Year

    Author Name. Project title, Year. Accessed: Month Day, Year. 6

  27. [35]

    Learning transferable visual models from natural language supervi- sion

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

    2021

  28. [36]

    Hierarchical text-conditional image gener- ation with clip latents

    Aditya Ramesh, Prafulla Dhariwal, Alex Nichol, Casey Chu, and Mark Chen. Hierarchical text-conditional image gener- ation with clip latents. arXiv preprint arXiv:2204.06125, 1 (2):3, 2022. 2, 3

    2022 arXiv

  29. [37]

    The sketchy database: learning to retrieve badly drawn bunnies

    Patsorn Sangkloy, Nathan Burnell, Cusuh Ham, and James Hays. The sketchy database: learning to retrieve badly drawn bunnies. ACM Transactions on Graphics (TOG), 35(4):1–12,

  30. [38]

    Laion-5b: An open large-scale dataset for training next generation image-text models

    Christoph Schuhmann, Romain Beaumont, Richard Vencu, Cade Gordon, Ross Wightman, Mehdi Cherti, Theo Coombes, Aarush Katta, Clayton Mullis, Mitchell Worts- man, et al. Laion-5b: An open large-scale dataset for training next generation image-text models. Advances in Neural In- f...

    2022

  31. [39]

    Learning to sketch with shortcut cycle consistency

    Jifei Song, Kaiyue Pang, Yi-Zhe Song, Tao Xiang, and Tim- othy M Hospedales. Learning to sketch with shortcut cycle consistency. In Proceedings of the IEEE conference on com- puter vision and pattern recognition , pages 801–810, 2018. 2, 3

    2018

  32. [40]

    Name your colour for the task: Artificially discover colour naming via colour quantisation transformer

    Shenghan Su, Lin Gu, Yue Yang, Zenghui Zhang, and Tat- suya Harada. Name your colour for the task: Artificially discover colour naming via colour quantisation transformer. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 12021–12031, 2023. 1

    2023

  33. [41]

    Clipasso: Semantically-aware object sketching

    Yael Vinker, Ehsan Pajouheshgar, Jessica Y Bo, Ro- man Christian Bachmann, Amit Haim Bermano, Daniel Cohen-Or, Amir Zamir, and Ariel Shamir. Clipasso: Semantically-aware object sketching. ACM Transactions on Graphics (TOG), 41(4):1–11, 2022. 2, 3, 4, 5, 7

    2022

  34. [42]

    Clipascene: Scene sketching with different types and levels of abstraction

    Yael Vinker, Yuval Alaluf, Daniel Cohen-Or, and Ariel Shamir. Clipascene: Scene sketching with different types and levels of abstraction. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 4146– 4156, 2023. 2, 3, 5

    2023

  35. [43]

    Learning robust global representations by penalizing local predictive power.Advances in Neural Information Pro- cessing Systems, 32, 2019

    Haohan Wang, Songwei Ge, Zachary Lipton, and Eric P Xing. Learning robust global representations by penalizing local predictive power.Advances in Neural Information Pro- cessing Systems, 32, 2019. 3

    2019

  36. [44]

    Diffsketching: Sketch control image synthesis with diffusion models

    Qiang Wang, Di Kong, Fengyin Lin, and Yonggang Qi. Diffsketching: Sketch control image synthesis with diffusion models. arXiv preprint arXiv:2305.18812, 2023. 3, 5

    2023 arXiv

  37. [45]

    Xdog: An extended difference-of-gaussians com- pendium including advanced image stylization

    Holger Winnem ¨oller, Jan Eric Kyprianidis, and Sven C Olsen. Xdog: An extended difference-of-gaussians com- pendium including advanced image stylization. Computers & Graphics, 36(6):740–753, 2012. 4

    2012

  38. [46]

    Svgdreamer: Text guided svg gener- ation with diffusion model

    Ximing Xing, Haitao Zhou, Chuang Wang, Jing Zhang, Dong Xu, and Qian Yu. Svgdreamer: Text guided svg gener- ation with diffusion model. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 4546–4555, 2024. 3

    2024

  39. [47]

    Drb-gan: A dynamic resblock generative adversar- ial network for artistic style transfer

    Wenju Xu, Chengjiang Long, Ruisheng Wang, and Guanghui Wang. Drb-gan: A dynamic resblock generative adversar- ial network for artistic style transfer. In Proceedings of the IEEE/CVF international conference on computer vision, pages 6383–6392, 2021. 3

    2021

  40. [48]

    On perceptual lossy compression: The cost of percep- tual reconstruction and an optimal training framework

    Zeyu Yan, Fei Wen, Rendong Ying, Chao Ma, and Peilin Liu. On perceptual lossy compression: The cost of percep- tual reconstruction and an optimal training framework. InIn- ternational Conference on Machine Learning, pages 11682– 11692. PMLR, 2021. 4

    2021

  41. [49]

    Im- proving fairness in image classification via sketching

    Ruichen Yao, Ziteng Cui, Xiaoxiao Li, and Lin Gu. Im- proving fairness in image classification via sketching. arXiv preprint arXiv:2211.00168, 2022. 3

    2022 arXiv

  42. [50]

    The evolution of color naming re- flects pressure for efficiency: Evidence from the recent past

    Noga Zaslavsky, Karee Garvin, Charles Kemp, Naftali Tishby, and Terry Regier. The evolution of color naming re- flects pressure for efficiency: Evidence from the recent past. Journal of Language Evolution, 7(2):184–199, 2022. 1

    2022

  43. [51]

    Inversion-based style transfer with diffusion models

    Yuxin Zhang, Nisha Huang, Fan Tang, Haibin Huang, Chongyang Ma, Weiming Dong, and Changsheng Xu. Inversion-based style transfer with diffusion models. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10146–10156, 2023. 2, 3

    2023

  44. [52]

    Unpaired image-to-image translation using cycle- consistent adversarial networks

    Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A Efros. Unpaired image-to-image translation using cycle- consistent adversarial networks. In Proceedings of the IEEE international conference on computer vision , pages 2223– 2232, 2017. 3

    2017

  45. [53]

    Sketchyscene: Richly-annotated scene sketches

    Changqing Zou, Qian Yu, Ruofei Du, Haoran Mo, Yi-Zhe Song, Tao Xiang, Chengying Gao, Baoquan Chen, and Hao Zhang. Sketchyscene: Richly-annotated scene sketches. In Proceedings of the european conference on computer vision (ECCV), pages 421–436, 2018. 3

    2018

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.