Pith. sign in

REVIEW 5 major objections 8 minor 43 references

CoCoNO: Attention Contrast-and-Complete for Initial Noise Optimization in Text-to-Image Synthesis

T0 review · 5 major / 8 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 CoCoNO claims that two attention-map defects explain most subject neglect and mixing in text-to-image models, and that optimizing the initial noise to fix them yields better aligned images than existing methods.

desk verdict A promising attention-based fix for subject mixing and neglect, but Eq. (3) quietly turns 'initial noise optimization' into latent refinement, so the paper needs a major reframing before the numbers are trustworthy. read the letter →

arxiv 2411.16783 v1 pith:AMGTRCT5 submitted 2024-11-25 cs.CV

Aravindan Sundaram , Ujjayan Pal , Abhimanyu Chauhan , Aishwarya Agarwal , Srikrishna Karanam This is my paper · ORCID
classification cs.CV
keywords initialnoiseoptimizationtext-to-imagesynthesisattentionmapssubjectneglectmixingdiffusionmodelslatentcontrastloss
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

CoCoNO claims that the remaining failures of text-to-image generation — subjects left out entirely (subject neglect) and subjects blended into one another (subject mixing) — have two specific attention-map causes: missing self-attention segments for some prompt tokens (attention neglect) and one token's cross-attention leaking into another token's self-attention segment (attention interference). The paper's contribution is a noise-optimization algorithm that fixes both causes at once by jointly reading self- and cross-attention maps on one-step denoised latents. Its two losses, attention complete and attention contrast, force each subject to own a distinct high-response segment and forbid cross-attention from spilling into other subjects' segments. On three benchmarks (T2I-CompBench, GenAI-Bench, and a custom Complex-Bench) and in a user study, CoCoNO reports larger image-text and text-text similarity scores than InitNO, CONFORM, Divide-and-Bind, Attend-and-Excite, and plain Stable Diffusion, with no model retraining.

What carries the argument

The central machinery is a pair of losses computed from aggregated self- and cross-attention maps of a one-step denoised latent. After sigmoid-softening and segmenting the self-attention map (Otsu thresholding plus connected components), a Hungarian assignment maps each of the n subject tokens to one of the n self-attention segments, padding with zero segments when the map has fewer than n segments. The attention complete loss maximizes the smallest diagonal intersection (Eq. 6), forcing every subject to have a present, high-response segment, while the attention contrast loss minimizes off-diagonal intersections (Eq. 5), preventing cross-attention interference. These losses, plus a KL term to stay near N(0,1), update the mean and covariance of the initial latent; the updated latent is denoised one step, and the loop repeats.

What would settle it

A concrete test: for prompts with two subjects, compute the one-step attention maps and check whether the CoCoNO segment conditions hold, then fully denoise and measure subject presence and mixing with a segmentation model such as Grounded-SAM. If a large fraction of latents whose one-step maps satisfy both conditions still produce missing or mixed subjects, the proxy assumption is false. A complementary check would perturb only the contrast term (or only the complete term) and observe whether the predicted mixing (or neglect) appears.

Watch

Extended reading notes

Core claim

The central claim, stated by the authors, is that attention neglect and attention interference — not merely high-response but non-overlapping attention maps — are the decisive causes of subject omission and subject mixing after initial-noise optimization, and that correcting them during initialization propagates to the final image. They demonstrate that when each subject token is assigned a dedicated self-attention segment (complete) and no token's cross-attention activates another subject's segment (contrast), the optimized latent denoises into images containing all prompt subjects with their own properties. This is established through quantitative comparisons on three benchmarks and a four-way user study in which CoCoNO is preferred 72.6% of the time.

Load-bearing premise

The method assumes that attention maps taken from a single one-step denoised latent reliably predict where subjects will end up in the fully denoised image; the paper motivates this with examples but does not quantify the correlation.

Editorial extensions

If this is right

  • CoCoNO requires no retraining: it optimizes only the initial noise's mean and covariance, so it is directly compatible with any latent diffusion backbone.
  • The attention complete loss creates self-attention segments that were entirely absent, directly targeting the case where a subject is omitted; the attention contrast loss prevents cross-attention leakage, targeting mixed properties.
  • Because each subject token is assigned to exactly one segment by the Hungarian step, the method naturally enforces a one-to-one subject-to-region mapping.
  • Reported inference time rises from 7.71s (Stable Diffusion) to 17.21s (CoCoNO) per 512x512 image, a modest add-on for test-time alignment gains.

Reading between the lines

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

  • The same contrast-complete principle could be applied to the cross-attention maps during the denoising trajectory itself, not only at initialization, which might correct misplacements that appear only in later sampling steps.
  • The hard Hungarian assignment assumes a fixed number of subject tokens; extending to variable-length prompts or modeling inter-subject relations (a limitation the authors name) may require a soft assignment or an auxiliary relation model.
  • The loss values themselves could serve as a diagnostic: a prompt that keeps the complete loss near one after optimization is likely structurally unrenderable by the backbone, giving users early feedback on prompt feasibility.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 8 minor

Summary. The paper proposes CoCoNO, a training-free method for optimizing the initial latent of Stable Diffusion v2.1 for text-to-image generation. Two attention-based losses are introduced: an attention complete loss, which encourages each subject token to own a distinct high-response segment in the self-attention map, and an attention contrast loss, which penalizes overlap between one subject's cross-attention map and another subject's self-attention segment. These losses are combined with a KL regularization term and minimized by updating the parameters of the latent distribution over repeated one-step denoising iterations. The authors evaluate on T2I-CompBench, GenAI-Bench, and a newly curated 20-prompt Complex-Bench, reporting higher VQAScore and text-text CLIP similarity than Stable Diffusion, InitNO, CONFORM, Divide-and-Bind, and Attend-and-Excite, along with a user study favoring CoCoNO.

Significance. If the empirical claims hold, CoCoNO would be a useful training-free contribution to improving subject coverage and reducing subject mixing in text-to-image generation. The problem identification is clear, and the two losses address a genuine limitation of prior attention-based methods. The paper has notable strengths: the central evaluation metrics (VQAScore and text-text CLIP similarity) are independent of the proposed losses, and the ablation in Table 1 shows that each loss contributes to the reported gain. However, the load-bearing update rule in Eq. (3) does not produce a valid Gaussian initial noise sample, and the headline comparisons lack error bars and significance tests. These issues are substantial but appear addressable, so the contribution remains potentially significant after revision.

major comments (5)
  1. [Section 3.3, Eq. (3)] The updated latent z'_T = mu' + sigma' z_{T-1} is not a sample of the Gaussian initial noise distribution that the method claims to optimize. Since z_{T-1} is the output of one-step denoising, sigma' z_{T-1} is a scaled, structured latent rather than an independent Gaussian sample; its variance is sigma'^2 Var(z_{T-1}), which is not controlled by the KL loss in Eq. (7). The KL term KL(N(mu, sigma^2) || N(0,1)) regularizes the parameters, not the actual distribution of z'_T, so the stated constraint is not enforced. This changes the interpretation from initial noise optimization to a form of iterative latent refinement and could make comparisons with noise-optimization baselines inequitable. Please either reparameterize with fresh noise epsilon ~ N(0,I) at each iteration, or revise the framing and the KL loss to match the actual distribution of the updated latent.
  2. [Section 3.2 and Figure 4] The core premise that attention maps computed from a single one-step denoised latent are causally predictive of final image composition is supported only by qualitative examples. If one-step attention maps do not reliably indicate where subjects will appear after full denoising, Eqs. (5)-(6) optimize an auxiliary quantity whose transfer to the final image is not guaranteed. The authors should quantify this relationship, for example by comparing segments obtained from one-step attention maps with final-image subject masks from Grounded-SAM across a sample of prompts and reporting agreement as a function of timestep or optimization iteration. This would also justify the choice of one-step denoising rather than a later timestep.
  3. [Figure 7 and Table 1] The headline quantitative results are averages over 64 generated images per prompt, with only 20 prompts for Complex-Bench, and no variance, confidence intervals, or significance tests are reported. Without these, the claim of significant improvement over baselines is not established; the differences in Figure 7 may be within seed-level noise. Please report per-prompt distributions, standard errors, and pairwise significance tests (for example bootstrap or paired tests), and specify the number of prompts per benchmark for the aggregated numbers in Table 1.
  4. [Equations (5)-(6)] The losses as written divide by the sum of A_s^i over its spatial dimensions. When a subject has a missing self-attention segment, the paper explicitly creates zero-element matrices for that segment, making the denominator zero and the loss expression undefined (0/0 for a missing diagonal segment) unless a numerical epsilon is added. Please define the normalization for missing segments, and clarify whether the assignment permutation is recomputed after zero segments are populated during optimization.
  5. [Section 3.3 and Appendix A.2] The method depends on several empirical choices without sensitivity analysis: lambda_1=1, lambda_2=1, lambda_3=500, alpha=16, beta=0.5, Adam learning rate 10^{-2}, restart thresholds, and the latent cache. Given that lambda_3 is three orders of magnitude larger than the attention losses and that the KL loss is miscalibrated under Eq. (3), it is important to show that the reported performance is not a knife-edge artifact of a particular weight combination. A sensitivity sweep over at least the loss weights and alpha/beta would substantially strengthen the claim.
minor comments (8)
  1. [Abstract] The abstract contains a duplicated word ('despite despite') and the title in the full text has a spacing error ('No ise'); these should be corrected.
  2. [Section 3.3] The notation 'P ⊗ C' is described as matrix multiplication, but the tensor-product symbol is commonly read as elementwise or Kronecker product; please use a standard matrix product notation such as P C.
  3. [Figure 5] The figure contains garbled text ('K VQK VQ') and the formulas for the loss computation and latent update are hard to read; please redraw it with clear notation.
  4. [Section 4.2] Table 1 does not state whether the reported scores are averaged over all three benchmarks or a subset; please state this explicitly and give the number of prompts used for each aggregated value.
  5. [Section 4.2, Table 2] The user study reports only the percentage of preferences; please also report the number of participants, the number of comparisons per participant, and inter-rater agreement.
  6. [Appendix A.3, Figure 10] The Grounded-SAM mask-count and overlap evaluation measures almost exactly what the proposed losses are designed to enforce, so it is not an independent evaluation metric; please present it as a diagnostic of the intended mechanism rather than as evidence that general text-image alignment improves.
  7. [References] Reference [2] lists only 'Tom B Brown' as the author; this should be 'Brown et al.' with the full author list.
  8. [Appendix A.2] The sentence 'from The subject tokens use during the latent optimization...' contains a grammatical error and should be rewritten.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: CoCoNO's attention losses are evaluated with external metrics, not by the loss definitions themselves.

full rationale

The paper's derivation chain optimizes (mu, sigma) using attention contrast and complete losses computed on one-step denoised latents, then generates images from the optimized starting latent. The central quantitative claims are backed by VQAScore, BLIP text-text similarity, and a user study, all of which are external to the losses and to the self-attention/cross-attention maps used in optimization; these do not reduce to the loss definitions. The Appendix A.3 Grounded-SAM evaluation measures mask counts and overlap on final generated images, which is conceptually aligned with the attention neglect/interference objectives but is computed on the final decoded image by an external detector, so it is not the same quantity as Eqs. (5)-(6) and is not forced by construction. The self-citation to [1] (A-STAR) is used only for the standard fact that cross-attention maps arise at each denoising step, not to justify the central claim, so it is not load-bearing. One technical concern (not circularity) is that Eq. (3) defines the updated latent as z'_T = mu' + sigma' z_{T-1}, where z_{T-1} is a one-step denoised latent, so the KL loss in Eq. (7) may be miscalibrated; however, this is a correctness/premise issue, not a reduction of the result to its own inputs. No step in the claimed derivation is equivalent by construction to the losses or to a fitted parameter, so the circularity score is 0.

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

No new physical entities are introduced. 'Attention neglect' and 'attention interference' are conceptual failure-mode labels, not entities with falsifiable handles of their own.

free parameters (4)
  • lambda_1, lambda_2, lambda_3 (loss weights) = 1, 1, 500
    Equation 7 weights for L_Acont, L_Acomp, L_KL; set empirically without sensitivity analysis.
  • alpha, beta (sigmoid scalars) = alpha=16, beta=0.5
    Section 3.3; controls sharpness of self-attention segmentation.
  • Adam learning rate = 1e-2
    Appendix A.2; learning rate for updating mu and sigma.
  • Optimization restart thresholds = Not specified
    Appendix A.2 imposes an upper limit on steps and loss-decrease thresholds but does not report their values, making the stopping rule a hidden parameter.
assumptions (4)
  • domain assumption One-step denoised latents yield attention maps informative of final image composition.
    Section 3.2 and Figure 4; the whole optimization uses these maps as the training signal.
  • domain assumption Self-attention PCA first principal component, binarized with Otsu and segmented with BFS, yields segments that correspond to subjects.
    Section 3.3 and Appendix A.2; if segmentation fails, the assignment and losses operate on the wrong regions.
  • standard math Hungarian assignment with cost C(i,j)=sum(As_i * Ac_j) correctly maps each subject to a segment.
    Section 3.3; a standard assignment algorithm, but its correctness here depends on the cost matrix being a faithful measure of subject-segment correspondence.
  • domain assumption The KL penalty with lambda_3=500 keeps the optimized latent close enough to N(0,1) to denoise in distribution.
    Section 3.3, Equation 7; prevents out-of-distribution latents but is not validated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CoCoNO: Attention Contrast-and-Complete for Initial Noise Optimization in Text-to-Image Synthesis." pith.science (2026). https://pith.science/paper/AMGTRCT5

@misc{pith2026241116783,
  author       = {Pith},
  title        = {Pith review of: CoCoNO: Attention Contrast-and-Complete for Initial Noise Optimization in Text-to-Image Synthesis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AMGTRCT5}},
  note         = {Machine review of arXiv:2411.16783}
}
read the original abstract

Despite recent advancements in text-to-image models, achieving semantically accurate images in text-to-image diffusion models is a persistent challenge. While existing initial latent optimization methods have demonstrated impressive performance, we identify two key limitations: (a) attention neglect, where the synthesized image omits certain subjects from the input prompt because they do not have a designated segment in the self-attention map despite despite having a high-response cross-attention, and (b) attention interference, where the generated image has mixed-up properties of multiple subjects because of a conflicting overlap between cross- and self-attention maps of different subjects. To address these limitations, we introduce CoCoNO, a new algorithm that optimizes the initial latent by leveraging the complementary information within self-attention and cross-attention maps. Our method introduces two new loss functions: the attention contrast loss, which minimizes undesirable overlap by ensuring each self-attention segment is exclusively linked to a specific subject's cross attention map, and the attention complete loss, which maximizes the activation within these segments to guarantee that each subject is fully and distinctly represented. Our approach operates within a noise optimization framework, avoiding the need to retrain base models. Through extensive experiments on multiple benchmarks, we demonstrate that CoCoNO significantly improves text-image alignment and outperforms the current state of the art.

Figures

Figures reproduced from arXiv: 2411.16783 by the authors.

Figure 1
Figure 1. We propose CoCoNO, a new algorithm for determining the best initial latent noise which when denoised with text-to-image [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Images demonstrating subject mixing and neglect. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Our proposed CoCoNO alleviates subject neglect and mixing by ensuring one high-response self-attention segment for each [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (9 more)
Figure 5
Figure 5. Figure 5: A visual illustration of proposed method. [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 4
Figure 4. Figure 4: Intermediate one-step denoised attention maps. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 6
Figure 6. Figure 6: Qualitative comparisons of CoCoNO with recent state-of-the-art methods. [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Average image-text and text-text similarities between the text prompts and the images generated by each method. [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 8
Figure 8. Figure 8: Qualitative ablation results. 4.2. Results Qualitative Comparison. We first present a qualitative comparison with our baselines in [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: Our proposed CoCoNO alleviates subject neglect and mixing by ensuring one high-response self-attention segment for each [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: Average number of distinct segments and corresponding overlap score for images generated by each method. [PITH_FULL_IMAGE:figures/full_fig_p012_10.png]
Figure 11
Figure 11. Figure 11: Additional Qualitative comparisons of CoCoNO with recent state-of-the-art methods. [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]
Figure 12
Figure 12. Figure 12: Additional Qualitative comparisons of CoCoNO with recent state-of-the-art methods. [PITH_FULL_IMAGE:figures/full_fig_p013_12.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 20 canonical work pages

  1. [1]

    A-star: Test-time attention segregation and retention for text-to-image synthesis

    Aishwarya Agarwal, Srikrishna Karanam, KJ Joseph, Apoorv Saxena, Koustava Goswami, and Balaji Vasan Srini- vasan. A-star: Test-time attention segregation and retention for text-to-image synthesis. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 2283– 2293, 2023. 2, 4

    work page 2023

  2. [2]

    Language models are few-shot learners

    Tom B Brown. Language models are few-shot learners. arXiv preprint ArXiv:2005.14165, 2020. 8, 9, 10

    arXiv 2005

  3. [3]

    Attend-and-excite: Attention-based se- mantic guidance for text-to-image diffusion models

    Hila Chefer, Yuval Alaluf, Yael Vinker, Lior Wolf, and Daniel Cohen-Or. Attend-and-excite: Attention-based se- mantic guidance for text-to-image diffusion models. ACM Transactions on Graphics (TOG), 42(4):1–10, 2023. 2, 4, 8, 9, 11

    work page 2023

  4. [4]

    Reno: Enhancing one-step text-to-image models through reward-based noise optimiza- tion

    Luca Eyring, Shyamgopal Karthik, Karsten Roth, Alexey Dosovitskiy, and Zeynep Akata. Reno: Enhancing one-step text-to-image models through reward-based noise optimiza- tion. arXiv preprint arXiv:2406.04312, 2024. 2

    arXiv 2024

  5. [5]

    Prompt optimizer of text-to-image diffusion models for abstract con- cept understanding

    Zezhong Fan, Xiaohan Li, Kaushiki Nag, Chenhao Fang, Topojoy Biswas, Jianpeng Xu, and Kannan Achan. Prompt optimizer of text-to-image diffusion models for abstract con- cept understanding. In Companion Proceedings of the ACM on Web Conference 2024, pages 1530–1537, 2024. 2, 4

    work page 2024

  6. [6]

    Initno: Boosting text-to-image diffu- sion models via initial noise optimization

    Xiefan Guo, Jinlin Liu, Miaomiao Cui, Jiankai Li, Hongyu Yang, and Di Huang. Initno: Boosting text-to-image diffu- sion models via initial noise optimization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 9380–9389, 2024. 2, 3, 4, 5, 6, 8, 9, 11

    work page 2024

  7. [7]

    Optimizing prompts for text-to-image generation

    Yaru Hao, Zewen Chi, Li Dong, and Furu Wei. Optimizing prompts for text-to-image generation. Advances in Neural Information Processing Systems, 36, 2024. 2, 4

    work page 2024

  8. [8]

    Prompt-to-prompt im- age editing with cross attention control

    Amir Hertz, Ron Mokady, Jay Tenenbaum, Kfir Aberman, Yael Pritch, and Daniel Cohen-Or. Prompt-to-prompt im- age editing with cross attention control. arXiv preprint arXiv:2208.01626, 2022. 2, 4

    arXiv 2022

Show all 43 references
  1. [9]

    Classifier-free diffusion guidance

    Jonathan Ho and Tim Salimans. Classifier-free diffusion guidance. arXiv preprint arXiv:2207.12598, 2022. 4

    2022 arXiv

  2. [10]

    Introvae: Introspective variational autoencoders for photo- 13 graphic image synthesis

    Huaibo Huang, Ran He, Zhenan Sun, Tieniu Tan, et al. Introvae: Introspective variational autoencoders for photo- 13 graphic image synthesis. Advances in neural information processing systems, 31, 2018. 4

    2018

  3. [11]

    T2i-compbench: A comprehensive bench- mark for open-world compositional text-to-image genera- tion

    Kaiyi Huang, Kaiyue Sun, Enze Xie, Zhenguo Li, and Xihui Liu. T2i-compbench: A comprehensive bench- mark for open-world compositional text-to-image genera- tion. Advances in Neural Information Processing Systems , 36:78723–78747, 2023. 8

    2023

  4. [12]

    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,

  5. [13]

    A style-based generator architecture for generative adversarial networks

    Tero Karras, Samuli Laine, and Timo Aila. A style-based generator architecture for generative adversarial networks. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 4401–4410, 2019. 4

    2019

  6. [14]

    Auto-encoding varia- tional bayes

    Diederik P Kingma and Max Welling. Auto-encoding varia- tional bayes. arXiv preprint arXiv:1312.6114, 2013. 4, 6

    2013 arXiv

  7. [15]

    Segment any- thing

    Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer White- head, Alexander C Berg, Wan-Yen Lo, et al. Segment any- thing. In Proceedings of the IEEE/CVF International Con- ference on Computer Vision, pages 4015–4026, 2023. 9, 11

    2023

  8. [16]

    The hungarian method for the assignment problem

    Harold W Kuhn. The hungarian method for the assignment problem. Naval research logistics quarterly , 2(1-2):83–97,

  9. [17]

    An algorithm for path connections and its applications

    Chin Yang Lee. An algorithm for path connections and its applications. IRE transactions on electronic computers, (3): 346–365, 1961. 9

    1961

  10. [18]

    Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation

    Junnan Li, Dongxu Li, Caiming Xiong, and Steven Hoi. Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation. In Interna- tional conference on machine learning, pages 12888–12900. PMLR, 2022. 8

    2022

  11. [19]

    Divide & bind your attention for improved generative seman- tic nursing

    Yumeng Li, Margret Keuper, Dan Zhang, and Anna Khoreva. Divide & bind your attention for improved generative seman- tic nursing. In 34th British Machine Vision Conference 2023, BMVC 2023, 2023. 8, 9, 11

    2023

  12. [20]

    Evaluating text-to-visual generation with image-to-text gen- eration

    Zhiqiu Lin, Deepak Pathak, Baiqi Li, Jiayao Li, Xide Xia, Graham Neubig, Pengchuan Zhang, and Deva Ramanan. Evaluating text-to-visual generation with image-to-text gen- eration. arXiv preprint arXiv:2404.01291, 2024. 8

    2024 arXiv

  13. [21]

    Grounding dino: Marrying dino with grounded pre-training for open-set object detection

    Shilong Liu, Zhaoyang Zeng, Tianhe Ren, Feng Li, Hao Zhang, Jie Yang, Chunyuan Li, Jianwei Yang, Hang Su, Jun Zhu, et al. Grounding dino: Marrying dino with grounded pre-training for open-set object detection. arXiv preprint arXiv:2303.05499, 2023. 9, 11

    2023 arXiv

  14. [22]

    Improving text- to-image consistency via automatic prompt optimization

    Oscar Ma ˜nas, Pietro Astolfi, Melissa Hall, Candace Ross, Jack Urbanek, Adina Williams, Aishwarya Agrawal, Adri- ana Romero-Soriano, and Michal Drozdzal. Improving text- to-image consistency via automatic prompt optimization. arXiv preprint arXiv:2403.17804, 2024. 2, 4

    2024 arXiv

  15. [23]

    Conform: Contrast is all you need for high- fidelity text-to-image diffusion models

    Tuna Han Salih Meral, Enis Simsar, Federico Tombari, and Pinar Yanardag. Conform: Contrast is all you need for high- fidelity text-to-image diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 9005–9014, 2024. 2, 8, 9, 11

    2024

  16. [24]

    Dynamic prompt optimizing for text- to-image generation

    Wenyi Mo, Tianyu Zhang, Yalong Bai, Bing Su, Ji-Rong Wen, and Qing Yang. Dynamic prompt optimizing for text- to-image generation. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 26627–26636, 2024. 2, 4

    2024

  17. [25]

    Semantic image synthesis with spatially-adaptive nor- malization

    Taesung Park, Ming-Yu Liu, Ting-Chun Wang, and Jun-Yan Zhu. Semantic image synthesis with spatially-adaptive nor- malization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2337–2346,

  18. [26]

    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

  19. [27]

    Exploring the limits of transfer learning with a unified text-to-text transformer

    Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, and Peter J Liu. Exploring the limits of transfer learning with a unified text-to-text transformer. Journal of machine learning research, 21(140):1–67, 2020. 4

    2020

  20. [28]

    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

    2022 arXiv

  21. [29]

    High-resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj ¨orn Ommer. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10684–10695, 2022. 4, 8, 9, 11

    2022

  22. [30]

    Photorealistic text-to-image diffusion models with deep language understanding

    Chitwan Saharia, William Chan, Saurabh Saxena, Lala Li, Jay Whang, Emily L Denton, Kamyar Ghasemipour, Raphael Gontijo Lopes, Burcu Karagol Ayan, Tim Salimans, et al. Photorealistic text-to-image diffusion models with deep language understanding. Advances in neural information...

    2022

  23. [31]

    Notes on kullback-leibler divergence and likelihood

    Jonathon Shlens. Notes on kullback-leibler divergence and likelihood. arXiv preprint arXiv:1404.2000, 2014. 6

    2000 arXiv

  24. [32]

    Df-gan: A simple and effec- tive baseline for text-to-image synthesis

    Ming Tao, Hao Tang, Fei Wu, Xiao-Yuan Jing, Bing-Kun Bao, and Changsheng Xu. Df-gan: A simple and effec- tive baseline for text-to-image synthesis. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 16515–16525, 2022. 4

    2022

  25. [33]

    Plug-and-play diffusion features for text-driven image-to-image translation

    Narek Tumanyan, Michal Geyer, Shai Bagon, and Tali Dekel. Plug-and-play diffusion features for text-driven image-to-image translation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 1921–1930, 2023. 4

    1921

  26. [34]

    Neural discrete representation learning

    Aaron Van Den Oord, Oriol Vinyals, et al. Neural discrete representation learning. Advances in neural information pro- cessing systems, 30, 2017. 4, 6

    2017

  27. [35]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszko- reit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in neural information processing systems, 30, 2017. 4

    2017

  28. [36]

    Investigating prompt engineering in diffusion models

    Sam Witteveen and Martin Andrews. Investigating prompt engineering in diffusion models. arXiv preprint arXiv:2211.15462, 2022. 2, 4 14

    2022 arXiv

  29. [37]

    Attngan: Fine- grained text to image generation with attentional generative adversarial networks

    Tao Xu, Pengchuan Zhang, Qiuyuan Huang, Han Zhang, Zhe Gan, Xiaolei Huang, and Xiaodong He. Attngan: Fine- grained text to image generation with attentional generative adversarial networks. In Proceedings of the IEEE conference on computer vision and pattern recognition , page...

    2018

  30. [38]

    Improving text-to-image synthesis us- ing contrastive learning

    Hui Ye, Xiulong Yang, Martin Takac, Rajshekhar Sunderra- man, and Shihao Ji. Improving text-to-image synthesis us- ing contrastive learning. arXiv preprint arXiv:2107.02423 ,

    arXiv

  31. [39]

    Image binarization using otsu threshold- ing algorithm

    Jamileh Yousefi. Image binarization using otsu threshold- ing algorithm. Ontario, Canada: University of Guelph , 10,

  32. [40]

    Cross-modal contrastive learning for text-to- image generation

    Han Zhang, Jing Yu Koh, Jason Baldridge, Honglak Lee, and Yinfei Yang. Cross-modal contrastive learning for text-to- image generation. In Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition , pages 833–842, 2021. 4

    2021

  33. [41]

    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

    2017

  34. [42]

    To- ward multimodal image-to-image translation

    Jun-Yan Zhu, Richard Zhang, Deepak Pathak, Trevor Dar- rell, Alexei A Efros, Oliver Wang, and Eli Shechtman. To- ward multimodal image-to-image translation. Advances in neural information processing systems, 30, 2017

    2017

  35. [43]

    Dm-gan: Dynamic memory generative adversarial networks for text- to-image synthesis

    Minfeng Zhu, Pingbo Pan, Wei Chen, and Yi Yang. Dm-gan: Dynamic memory generative adversarial networks for text- to-image synthesis. In Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition , pages 5802–5810, 2019. 4 15

    2019

Pith tools

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