Pith. sign in

REVIEW 3 major objections 5 minor 116 references

RichControl: Structure- and Appearance-Rich Training-Free Spatial Control for Text-to-Image Generation

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Spatial control improves when condition features skip ahead to timestep 600.

desk verdict Decoupled injection schedule is a real win on SDXL, but the FLUX generality claim is undercut by their own supplementary note. read the letter →

arxiv 2507.02792 v5 pith:5WY5CHS7 submitted 2025-07-03 cs.CV

Lexi Pang , Liheng Zhang , Hang Ye , Xiaoxuan Ma , Yizhou Wang This is my paper
classification cs.CV
keywords text-to-imagegenerationtraining-freespatialcontrolfeatureinjectionschedulediffusionmodelsstructurepreservationappearancequalityrestartrefinement
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

This paper argues that training-free spatial control of text-to-image diffusion models fails when the condition image's features are injected at the same denoising step as the output image. The authors identify a temporal trade-off: as diffusion progresses, condition features become structurally richer but increasingly out-of-distribution relative to natural image features. The central proposal is a decoupled injection schedule, sampling condition features from a fixed middle timestep ($g(t)=600$) rather than synchronously, which balances structure preservation and appearance quality. The paper adds a restart refinement schedule and an appearance-rich prompting step to suppress artifacts, and reports the strongest results among the compared methods across seven condition types.

What carries the argument

The load-bearing object is the injection schedule $g(t)$ used in the substitution where the generation branch's feature $f_{l,t}$ becomes $f_{l,g(t)}^{struct}$ and its attention map $A_{l,t}$ becomes $A_{l,g(t)}^{struct}$, decoupling when condition features are extracted from the denoising process. A restart refinement schedule then iteratively adds noise and re-denoises within a fixed interval to wash out artifacts, and appearance-rich prompting uses a multimodal language model to rewrite the prompt so the appearance image matches the condition image's semantics.

What would settle it

A systematic grid over constant schedules $g(t)=C$ on the same SDXL setup that shows synchronous injection $g(t)=t$ matching or beating $C=600$ on both structural and appearance metrics, or a full-dimensional feature analysis that reverses the KL/self-similarity trade-off, would falsify the central claim.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the sampling schedule of condition features is a previously unexplored control variable, and the correct setting is asynchronous: inject features and attention maps extracted at a medium timestep, not the current denoising timestep. This is supported by measurements showing that structural self-similarity between condition and natural features improves as noise decreases while KL divergence between their distributions widens, so no single synchronized timestep can simultaneously maximize structure and domain alignment. The paper claims that any schedule whose values cluster around medium timesteps performs well regardless of functional form, and that the constant schedule $g(t)=600$ with caching is simple, effective, and computationally efficient.

Load-bearing premise

The central analysis assumes that projecting diffusion features onto one PCA component and measuring KL divergence with Gaussian density estimates faithfully represents the domain gap that actually hurts generation quality; if that proxy is misleading, the schedule conclusions may not transfer.

Editorial extensions

If this is right

  • Injection-based training-free control gains a new tuning axis: the injection timestep can be chosen independently of the denoising timestep.
  • Because condition features are extracted once and cached, the decoupled schedule can be cheaper than synchronous injection while improving quality.
  • The framework can be dropped into existing methods such as FreeControl and into transformer-based backbones such as FLUX without retraining.
  • Multiple spatial conditions can be composed in a single generation, which the paper demonstrates with prompts and several condition images.
  • The paper claims its training-free method surpasses training-based controllable generation in structural fidelity on challenging conditions, including on transformer architectures.

Reading between the lines

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

  • The optimal constant $C$ is likely model- and schedule-dependent; a natural extension is to select it automatically per backbone or per condition modality instead of fixing it at 600.
  • If the middle-stage dominance reflects a general property of diffusion feature distributions, similar asynchronous schedules could improve other attention-based editing or style-transfer methods that currently inject synchronously.
  • A testable extension would be to replace the PCA and Gaussian-density proxy with full-dimensional feature statistics to see whether the same trade-off curve is recovered.
  • Because the prompt-rewriting step relies on a multimodal language model, its behavior on rare or abstract conditions may inherit that model's biases; a version using only visual cues would isolate the contribution of the schedule itself.
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

3 major / 5 minor

Summary. The paper proposes RichControl, a training-free framework for conditional text-to-image generation. Its central idea is to decouple the condition-feature injection timestep from the denoising process via a schedule g(t), and, after an ablation over constant, linear, and nonlinear schedules, to recommend a constant schedule g(t)=600 on SDXL. The framework combines this structure-rich injection with a restart refinement schedule and an appearance-rich prompting stage, and reports improved structural alignment, prompt alignment, and user preference against training-free and some training-based baselines. It also demonstrates plug-in use with FreeControl and an extension to the DiT-based FLUX backbone.

Significance. If the central claims are supported, the paper opens a practical new tuning axis for training-free controllable generation, with a simple constant schedule that is easy to implement and cache, and it provides a useful analysis of feature-domain mismatch. The component design is modular, the qualitative coverage spans many condition types, and the plug-in demonstration with FreeControl is a strength. However, the load-bearing schedule conclusion is currently validated only on a small, self-constructed dataset, and the FLUX extension abandons the proposed schedule, so the scope of the contribution is narrower than the abstract states.

major comments (3)
  1. [Supp. E.1, §5.4, Fig. 7(b)] The FLUX experiments do not test the decoupled schedule. Supp. E.1 states that for FLUX "a synchronous schedule g(t)=t performs slightly better than the constant schedule g(t)=C, so we adopt the former," while the main text and Fig. 7(b) claim the framework "generalizes across architectures." The comparison with Qwen-Image-Edit and FLUX Kontext in Fig. 7(b) therefore provides no evidence for the paper's central mechanism on DiT backbones; it shows only that the pipeline with synchronous injection, restart refinement, and appearance-rich prompting can be run on FLUX. This is an internal inconsistency between the abstract's generality claim and the implementation used for the FLUX results. Please add a FLUX ablation over g(t) (including g(t)=C) or explicitly limit the generalization claim to the UNet/SDXL setting.
  2. [§5.3, §5.1, Tab. 2, Tab. 4, Supp. E.6] The central schedule parameters are selected and reported on the same evaluation data. C=600 and tau=400 are fixed by an ablation (Fig. 6, Fig. 16, Supp. E.6) over the same dataset used for the final quantitative comparisons in Tab. 4 and the user study, and the same applies to N=3 for restart refinement. The text says all experiments were repeated three times, yet no standard deviations, confidence intervals, or significance tests are reported anywhere in the tables. The difference in Self-Sim between Ours (0.096) and Ctrl-X (0.104) in Tab. 4, for instance, is impossible to interpret without variance and could easily be noise on roughly 130 condition-prompt pairs. The paper should either hold out a validation set for hyperparameter selection or report error bars and per-condition breakdowns, and it should avoid describing the empirical optimum as a "principled" model-derived result.
  3. [§3, Supp. C.1] The motivational analysis is a proxy whose link to generation quality is not established. Eq. (7) measures KL divergence between Gaussian KDEs fitted to the first PCA component of diffusion features; both the PCA projection and the 1D density estimate are substantial reductions of the diffusion feature space, and no evidence is given that this scalar tracks structure preservation or visual quality under injection. Similarly, the self-similarity curve in Fig. 2 is computed on DINO keys of images, not on the injected condition features. To make the claim that the decoupled schedule is "guided by" this analysis, please show a quantitative correspondence between the proxy metrics and the downstream Self-sim/ImageReward/HPSv2 values across the schedules in Fig. 6, or present the analysis as purely illustrative.
minor comments (5)
  1. [Fig. 6] The schedule labels A-I are not defined in the caption or in the surrounding text; please add a table or legend specifying the functional form of each schedule.
  2. [§5.1, Supp. E.5] The metric called "Condition LPIPS" is described in the main text as appearance quality, but Supp. E.5 defines it as perceptual deviation from the condition image, with higher being better; please make this direction explicit in the main text.
  3. [Supp. E.1] The FLUX implementation differs from the main SRI formulation by injecting only self-attention query matrices and by using different layers and timestep thresholds; please state explicitly how this relates to the feature and attention substitution described in §4.1.
  4. [Tab. 1, Supp. E.3] The user study reports a single preference rate of 56.25% without confidence intervals or per-case agreement; please report the number of votes per method and a variance estimate.
  5. [Supp. E.2] The evaluation dataset is self-constructed and small, but the paper does not state whether it will be released; please add an availability statement, since the quantitative claims rest entirely on this dataset.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the schedule constant is openly empirical, and the DiT inconsistency is a scope gap, not a definitional reduction.

full rationale

The claimed contribution is an empirical schedule design, not a first-principles derivation. Section 3's KL and self-similarity analysis motivates a mid-timestep injection, but it does not formally imply the specific constant C=600; that constant is selected through ablations reported in Sec. 5.3 and Supp. F.4, which explicitly state that "Empirically, C=600 achieves the best balance between appearance fidelity and structure control." Because the paper openly reports this as an empirical choice rather than relabeling a fitted constant as an independent prediction, the fitted-input-as-prediction pattern does not apply. The evaluation metrics (DINO Self-sim, CLIP, LPIPS, DreamSim, ImageReward, HPSv2) are external to the method's construction, so the comparisons are not equivalent to the method's own inputs by construction. The DiT extension in Supp. E.1 adopts a synchronous schedule for FLUX because it "performs slightly better" than the constant schedule; this weakens the claimed architecture-generality and is an internal-consistency/scope concern, but it is not a circular reduction of the conclusion to its inputs. The Limitations section likewise acknowledges that a theoretical interpretation is future work, consistent with the schedule being empirically tuned. No load-bearing self-citation, uniqueness import, or ansatz-via-citation pattern appears. Overall, the derivation chain is self-contained and genuinely empirical, so no significant circularity is found.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The central claim depends on a small set of hand-tuned scalar parameters (C, tau, N) that are chosen on the evaluation dataset. The empirical claims about a structure/quality trade-off in feature space rely on proxy metrics (KL on PCA projections) rather than a theoretical account. No new entities are introduced.

free parameters (7)
  • Injection timestep C = 600
    Chosen by ablation on the constructed dataset; the paper notes lower C causes leakage and higher C weakens structure.
  • Injection start tau = 400
    Hyperparameter controlling when injection begins; set without error analysis.
  • Restart iterations N = 3
    Ablated between 1, 3, and 5; 3 chosen as a balance of quality and cost.
  • Restart noise levels sigma_tmin/sigma_tmax = 1.0 / 2.0
    Set by hand; no sensitivity analysis reported.
  • Dilation and unsharp masking parameters = wmin=25, wmax=50, ke=10, gamma=50, r=3
    Empirical preprocessing choices described as beneficial without quantitative support.
  • FLUX guidance strength s = 6.5
    Standard heuristic for FLUX-dev; not analyzed.
  • FLUX injection layers = layers 25-38
    Likely tuned; rationale not given.
assumptions (4)
  • domain assumption The diffusion features of condition and natural images follow a distributional domain gap that is well captured by KL divergence and DINO self-similarity proxies.
    Motivates the decoupled schedule. The proxy analysis is presented as supporting evidence only, with no direct proof that this trade-off governs generation quality.
  • domain assumption SDXL self-attention features and attention maps are sufficient for structure control via substitution.
    Borrowed from Ctrl-X and PnP; accepted as background for training-free injection.
  • domain assumption GPT-4o semantic extraction improves appearance transfer without harming fidelity.
    The ARP module depends on an LLM correctly enriching prompts; no robustness analysis is given.
  • domain assumption The scheduling insights transfer from SDXL to FLUX.
    The paper reports that a synchronous schedule is preferred on FLUX, which contradicts the constant schedule finding, so the transfer is not guaranteed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of RichControl: Structure- and Appearance-Rich Training-Free Spatial Control for Text-to-Image Generation." pith.science (2026). https://pith.science/paper/5WY5CHS7

@misc{pith2026250702792,
  author       = {Pith},
  title        = {Pith review of: RichControl: Structure- and Appearance-Rich Training-Free Spatial Control for Text-to-Image Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5WY5CHS7}},
  note         = {Machine review of arXiv:2507.02792}
}
read the original abstract

Text-to-image (T2I) diffusion models have shown remarkable success in generating high-quality images from text prompts. Recent efforts extend these models to incorporate conditional images (e.g., canny edge) for fine-grained spatial control. Among them, feature injection methods have emerged as a training-free alternative to traditional fine-tuning-based approaches. However, they often suffer from structural misalignment, condition leakage, and visual artifacts, especially when the condition image diverges significantly from natural RGB distributions. Through an analysis of existing methods, we identify a key limitation: the sampling schedule of condition features, previously unexplored, fails to account for the evolving interplay between structure preservation and domain alignment throughout diffusion steps. Inspired by this observation, we propose a flexible training-free framework that decouples the sampling schedule of condition features from the denoising process, and systematically investigate the spectrum of feature injection schedules to achieve a better balance between structural alignment and appearance quality. We further enhance the sampling process by introducing a restart refinement schedule, and improve the visual quality with an appearance-rich prompting strategy. Together, these designs enable training-free controllable generation that is both structure-rich and appearance-rich. Extensive experiments demonstrate that our method achieves state-of-the-art performance under complex and diverse conditions. Owing to its generality, our framework naturally supports compositional conditional generation and generalizes across architectures in a plug-and-play manner, from UNet-based diffusion models to modern DiT backbones such as FLUX.

Figures

Figures reproduced from arXiv: 2507.02792 by the authors.

Figure 1
Figure 1. We propose a training-free framework that enables high-quality spatial control for pretrained text-to-image diffusion models under arbitrary spatial conditions. (Left) Our method addresses key limitations of prior training-free approaches to achieve more robust control; (Top right) can handle compositional generation with multiple conditions and complex prompts; and (Bottom right) can be flexibly extended to DiT-bas… view at source ↗
Figure 2
Figure 2. The evolving curves of KL divergence and L2 distance of self-similarity matrices across diffusion timesteps. Analysis. To validate this hypothesis, we quantitatively ana￾lyze features from 100 pairs of condition images across five common modalities (see Supp. Sec. C for more details). As shown by the orange curve in [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Visualizing diffusion features extracted from the condition and natural images at various timesteps. We also conduct principal component analysis and visu￾alize the diffusion features in [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (17 more)
Figure 4
Figure 4. Figure 4: Method overview. Given a condition image I struct and a prompt P, our method generates an output image I, aligning semantically with P while preserving the structure of I struct. Our framework consists of three key components. (i) The Structure-Rich Injection (SRI) mod…
Figure 5
Figure 5. Figure 5: Our method achieves SOTA performance among all training-free methods; in the radar chart, greater distance from the center indicates superior results. 5. Experiments 5.1. Setup Dataset. We construct our evaluation dataset based on datasets from prior work [49, 55]. Spe…
Figure 6
Figure 6. Figure 6: Ablation of SRI schedules. We report quantitative (Top) and qualitative (Bottom) results for different injection schedules. 5.2. Comparison with State-of-the-Art (SOTA) Analysis. Figs. 5 and 19 present quantitative and qualita￾tive comparisons between our method and th…
Figure 7
Figure 7. Figure 7: Flexibility of our framework. (a) As a plug-in, our method enhances FreeControl [55] with stronger structure preservation and visual quality. (b) Our method can be flexibly extended to the DiT-based FLUX backbone [40]. (c) Our method can handle compositional conditions…
Figure 8
Figure 8. Figure 8: Ablation of Appearance-Rich Prompting (ARP). See [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: Ablation of Restart Refinement (RR). See [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]
Figure 10
Figure 10. Figure 10: Fourier analysis of noisy latents under canny edge and depth map conditions. (Left) Average L2 distance between natural and condition image DFT spectra over timesteps. Subfigures (a)–(d) and (e)–(h) show the DFT spectra of four randomly selected images for both condit…
Figure 11
Figure 11. Figure 11: Average L2 distance between natural and condition image DFT spectra over diffusion timesteps. Results are aver￾aged over all five conditions. To further investigate how frequency components evolve through the diffusion process, we conduct a detailed analysis on two re…
Figure 12
Figure 12. Figure 12: Illustration of the Appearance-Rich Prompting (ARP) module. Given the original text prompt P, our module derives an appearance-rich prompt P app by integrating semantic information from the condition image I struct . semantic information, producing an enhanced appeara…
Figure 13
Figure 13. Figure 13: Screenshot of the user study interface. Participants are presented with the inputs and asked to select the best result from four randomly shuffled candidates. Notably, these metrics jointly assess both structure preserva￾tion (e.g., DINO self-similarity [81], DreamSim…
Figure 14
Figure 14. Figure 14: Additional ablation of restart iterations N. Setting N = 1 is not adequate for suppressing visual artifacts, and both N = 3 and N = 5 yield high-quality outputs. Consequently, we set N = 3 for optimal visual quality and computational efficiency. F.5. Additional Plug-i…
Figure 15
Figure 15. Figure 15: , our approach consistently improves FreeControl on all three metrics, demonstrating better structure alignment and appearance quality, further highlighting the flexibility of our framework [PITH_FULL_IMAGE:figures/full_fig_p020_15.png]
Figure 16
Figure 16. Figure 16: Additional ablation of Structure-Rich Injection (SRI). For asynchronous injection g(t) = C, lower C suffers from conditional leakage, while higher values improve appearance fidelity at the cost of structural control. The optimal trade-off is achieved at C = 600, outpe…
Figure 17
Figure 17. Figure 17: Additional ablation of Appearance-Rich Prompting (ARP). This module improves semantic alignment with the condition image by adapting prompts to better capture key visual attributes, thereby mitigating incorrect appearance transfers and reducing artifacts. A man dancin…
Figure 18
Figure 18. Figure 18: Additional ablation of Restart Refinement (RR). This strategy significantly mitigates condition leakage and appearance artifacts, improving generation quality while maintaining structural alignment [PITH_FULL_IMAGE:figures/full_fig_p022_18.png]
Figure 19
Figure 19. Figure 19: Qualitative comparison with existing methods [PITH_FULL_IMAGE:figures/full_fig_p023_19.png]
Figure 20
Figure 20. Figure 20: Qualitative results for more control conditions [PITH_FULL_IMAGE:figures/full_fig_p024_20.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

116 extracted references · 74 canonical work pages

  1. [1]

    Gpt-4 technical report.arXiv preprint arXiv:2303.08774,

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ah- mad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. Gpt-4 technical report.arXiv preprint arXiv:2303.08774,

    arXiv

  2. [2]

    Cross-image attention for zero- shot appearance transfer

    Yuval Alaluf, Daniel Garibi, Or Patashnik, Hadar Averbuch- Elor, and Daniel Cohen-Or. Cross-image attention for zero- shot appearance transfer. InACM SIGGRAPH Conference on Computer Graphics and Interactive Techniques (SIG- GRAPH), 2024. 3

    2024

  3. [3]

    Break-a-scene: Extracting multiple concepts from a single image

    Omri Avrahami, Kfir Aberman, Ohad Fried, Daniel Cohen- Or, and Dani Lischinski. Break-a-scene: Extracting multiple concepts from a single image. InACM SIGGRAPH Asia Conference on Computer Graphics and Interactive Tech- niques (SIGGRAPH Asia), 2023. 14

    2023

  4. [4]

    Spatext: Spatio-textual representation for con- trollable image generation

    Omri Avrahami, Thomas Hayes, Oran Gafni, Sonal Gupta, Yaniv Taigman, Devi Parikh, Dani Lischinski, Ohad Fried, and Xi Yin. Spatext: Spatio-textual representation for con- trollable image generation. InProceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2023. 14

    2023

  5. [5]

    Stable flow: Vital layers for training-free image editing

    Omri Avrahami, Or Patashnik, Ohad Fried, Egor Nemchi- nov, Kfir Aberman, Dani Lischinski, and Daniel Cohen-Or. Stable flow: Vital layers for training-free image editing. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition (CVPR), pages 7877–7888,

  6. [6]

    Universal guidance for diffusion models

    Arpit Bansal, Hong-Min Chu, Avi Schwarzschild, Soumyadip Sengupta, Micah Goldblum, Jonas Geiping, and Tom Goldstein. Universal guidance for diffusion models. InProceedings of International Conference on Learning Representations (ICLR), 2023. 3, 14

    2023

  7. [7]

    In- structpix2pix: Learning to follow image editing instructions

    Tim Brooks, Aleksander Holynski, and Alexei A Efros. In- structpix2pix: Learning to follow image editing instructions. InProceedings of Conference on Computer Vision and Pat- tern Recognition (CVPR), 2023. 14

    2023

  8. [8]

    Masactrl: Tuning-free mutual self-attention control for consistent image synthesis and editing

    Mingdeng Cao, Xintao Wang, Zhongang Qi, Ying Shan, Xiaohu Qie, and Yinqiang Zheng. Masactrl: Tuning-free mutual self-attention control for consistent image synthesis and editing. InProceedings of International Conference on Computer Vision (ICCV), 2023. 3

    2023

Show all 116 references
  1. [9]

    Z. Cao, G. Hidalgo Martinez, T. Simon, S. Wei, and Y . A. Sheikh. Openpose: Realtime multi-person 2d pose estima- tion using part affinity fields.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019. 14

    2019

  2. [10]

    Emerg- ing properties in self-supervised vision transformers

    Mathilde Caron, Hugo Touvron, Ishan Misra, Herv´e J´egou, Julien Mairal, Piotr Bojanowski, and Armand Joulin. Emerg- ing properties in self-supervised vision transformers. InPro- ceedings of International Conference on Computer Vision (ICCV), 2021. 5, 15

    2021

  3. [11]

    Training-free layout control with cross-attention guidance

    Minghao Chen, Iro Laina, and Andrea Vedaldi. Training-free layout control with cross-attention guidance. InProceedings of Winter Conference on Applications of Computer Vision (WACV), 2024. 3

    2024

  4. [12]

    Unireal: Universal image generation and editing via learn- ing real-world dynamics

    Xi Chen, Zhifei Zhang, He Zhang, Yuqian Zhou, Soo Ye Kim, Qing Liu, Yijun Li, Jianming Zhang, Nanxuan Zhao, Yilin Wang, Hui Ding, Zhe Lin, and Hengshuang Zhao. Unireal: Universal image generation and editing via learn- ing real-world dynamics. InProceedings of the IEEE/CVF Con...

    2025

  5. [13]

    Style in- jection in diffusion: A training-free approach for adapting large-scale diffusion models for style transfer

    Jiwoo Chung, Sangeek Hyun, and Jae-Pil Heo. Style in- jection in diffusion: A training-free approach for adapting large-scale diffusion models for style transfer. InProceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 8795–8805, 2024. 3

    2024

  6. [14]

    Diffedit: Diffusion-based semantic image editing with mask guidance

    Guillaume Couairon, Jakob Verbeek, Holger Schwenk, and Matthieu Cord. Diffedit: Diffusion-based semantic image editing with mask guidance. InProceedings of International Conference on Learning Representations (ICLR), 2023. 3

    2023

  7. [15]

    Fluxs- pace: Disentangled semantic editing in rectified flow trans- formers, 2024

    Yusuf Dalva, Kavana Venkatesh, and Pinar Yanardag. Fluxs- pace: Disentangled semantic editing in rectified flow trans- formers, 2024. 3

    2024

  8. [16]

    Freecustom: Tuning- free customized image generation for multi-concept compo- sition

    Ganggui Ding, Canyu Zhao, Wen Wang, Zhen Yang, Zide Liu, Hao Chen, and Chunhua Shen. Freecustom: Tuning- free customized image generation for multi-concept compo- sition. InProceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2024. 3

    2024

  9. [17]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Syl- vain Gelly, Jakob Uszkoreit, and Neil Houlsby. An image is worth 16x16 words: Transformers for image recognition ...

    2021

  10. [18]

    Efros, and Aleksander Holynski

    Dave Epstein, Allan Jabri, Ben Poole, Alexei A. Efros, and Aleksander Holynski. Diffusion self-guidance for control- lable image generation. InProceedings of Advances in Neural Information Processing Systems (NeurIPS), 2023. 3, 14

    2023

  11. [19]

    Scaling rec- tified flow transformers for high-resolution image synthesis

    Patrick Esser, Sumith Kulal, Andreas Blattmann, Rahim Entezari, Jonas M ¨uller, Harry Saini, Yam Levi, Dominik Lorenz, Axel Sauer, Frederic Boesel, Dustin Podell, Tim Dockhorn, Zion English, and Robin Rombach. Scaling rec- tified flow transformers for high-resolution image syn...

    2024

  12. [20]

    Personalize anything for free with diffusion trans- former.arXiv preprint arXiv:2503.12590, 2025

    Haoran Feng, Zehuan Huang, Lin Li, Hairong Lv, and Lu Sheng. Personalize anything for free with diffusion trans- former.arXiv preprint arXiv:2503.12590, 2025. 3

    2025 arXiv

  13. [21]

    Dit4edit: Diffusion transformer for image editing.Proceedings of the AAAI Conference on Artificial Intelligence, 39(3):2969– 2977, 2025

    Kunyu Feng, Yue Ma, Bingyuan Wang, Chenyang Qi, Haozhe Chen, Qifeng Chen, and Zeyu Wang. Dit4edit: Diffusion transformer for image editing.Proceedings of the AAAI Conference on Artificial Intelligence, 39(3):2969– 2977, 2025. 3

    2025

  14. [22]

    Dream- sim: Learning new dimensions of human visual similarity using synthetic data

    Stephanie Fu, Netanel Tamir, Shobhita Sundaram, Lucy Chai, Richard Zhang, Tali Dekel, and Phillip Isola. Dream- sim: Learning new dimensions of human visual similarity using synthetic data. InProceedings of Advances in Neural Information Processing Systems (NeurIPS), 2023. 5, 19

    2023

  15. [23]

    An image is worth one word: Personalizing text-to-image gener- ation using textual inversion

    Rinon Gal, Yuval Alaluf, Yuval Atzmon, Or Patashnik, Amit H Bermano, Gal Chechik, and Daniel Cohen-Or. An image is worth one word: Personalizing text-to-image gener- ation using textual inversion. InProceedings of International Conference on Learning Representations (ICLR), 2023. 14

    2023

  16. [24]

    Instructdiffusion: A generalist modeling interface for vision tasks

    Zigang Geng, Binxin Yang, Tiankai Hang, Chen Li, Shuyang Gu, Ting Zhang, Jianmin Bao, Zheng Zhang, Houqiang Li, Han Hu, Dong Chen, and Baining Guo. Instructdiffusion: A generalist modeling interface for vision tasks. InProceed- ings of the IEEE/CVF Conference on Computer Visio...

    2024

  17. [25]

    Eye-for-an-eye: Appearance transfer with semantic correspondence in diffusion models, 2024

    Sooyeon Go, Kyungmook Choi, Minjung Shin, and Youngjung Uh. Eye-for-an-eye: Appearance transfer with semantic correspondence in diffusion models, 2024. 3

    2024

  18. [26]

    Pair diffusion: A comprehensive multimodal object-level image editor

    Vidit Goel, Elia Peruzzo, Yifan Jiang, Dejia Xu, Xingqian Xu, Nicu Sebe, Trevor Darrell, Zhangyang Wang, and Humphrey Shi. Pair diffusion: A comprehensive multimodal object-level image editor. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition ...

    2024

  19. [27]

    Ace: All-round creator and editor following instructions via diffu- sion transformer, 2024

    Zhen Han, Zeyinzi Jiang, Yulin Pan, Jingfeng Zhang, Chao- jie Mao, Chenwei Xie, Yu Liu, and Jingren Zhou. Ace: All-round creator and editor following instructions via diffu- sion transformer, 2024. 14

    2024

  20. [28]

    Prompt-to-prompt image editing with cross attention control

    Amir Hertz, Ron Mokady, Jay Tenenbaum, Kfir Aberman, Yael Pritch, and Daniel Cohen-Or. Prompt-to-prompt image editing with cross attention control. InProceedings of Inter- national Conference on Learning Representations (ICLR),

  21. [29]

    Classifier-free diffusion guidance

    Jonathan Ho and Tim Salimans. Classifier-free diffusion guidance. InNeurIPS 2021 Workshop on Deep Generative Models and Downstream Applications, 2021. 2, 14

    2021

  22. [30]

    Denoising dif- fusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising dif- fusion probabilistic models. InProceedings of Advances in Neural Information Processing Systems (NeurIPS), 2020. 14

    2020

  23. [31]

    Cascaded diffu- sion models for high fidelity image generation.Journal of Machine Learning Research (JMLR), 2022

    Jonathan Ho, Chitwan Saharia, William Chan, David J Fleet, Mohammad Norouzi, and Tim Salimans. Cascaded diffu- sion models for high fidelity image generation.Journal of Machine Learning Research (JMLR), 2022. 2

    2022

  24. [32]

    Anchor token matching: Im- plicit structure locking for training-free ar image editing

    Taihang Hu, Linxuan Li, Kai Wang, Yaxing Wang, Jian Yang, and Ming-Ming Cheng. Anchor token matching: Im- plicit structure locking for training-free ar image editing. InProceedings of International Conference on Computer Vision (ICCV), 2025. 3

    2025

  25. [33]

    Attenst: A training-free attention-driven style trans- fer framework with pre-trained diffusion models, 2025

    Bo Huang, Wenlun Xu, Qizhuo Han, Haodong Jing, and Ying Li. Attenst: A training-free attention-driven style trans- fer framework with pre-trained diffusion models, 2025. 3

    2025

  26. [34]

    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. InProceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2017. 14

    2017

  27. [35]

    Yueru Jia, Aosong Cheng, Yuhui Yuan, Chuke Wang, Ji Li, Huizhu Jia, and Shanghang Zhang. Designedit: Unify spatial-aware image editing via training-free inpainting with a multi-layered latent diffusion framework.Proceedings of the AAAI Conference on Artificial Intelligence, 39...

    2025

  28. [36]

    Elucidating the design space of diffusion-based generative models

    Tero Karras, Miika Aittala, Timo Aila, and Samuli Laine. Elucidating the design space of diffusion-based generative models. InProceedings of Advances in Neural Information Processing Systems (NeurIPS), 2022. 2

    2022

  29. [37]

    Dif- fusionclip: Text-guided diffusion models for robust image manipulation

    Gwanghyun Kim, Taesung Kwon, and Jong Chul Ye. Dif- fusionclip: Text-guided diffusion models for robust image manipulation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 2426–2435, 2022. 14

    2022

  30. [38]

    Dense text-to-image generation with attention modulation

    Yunji Kim, Jiyoung Lee, Jin-Hwa Kim, Jung-Woo Ha, and Jun-Yan Zhu. Dense text-to-image generation with attention modulation. InProceedings of Conference on Computer Vision and Pattern Recognition (CVPR), pages 7701–7711,

  31. [39]

    Diffusion-based image translation using disentangled style and content representa- tion

    Gihyun Kwon and Jong Chul Ye. Diffusion-based image translation using disentangled style and content representa- tion. InProceedings of International Conference on Learn- ing Representations (ICLR), 2023. 3

    2023

  32. [40]

    Black Forest Labs. Flux. https://github.com/ black-forest-labs/flux, 2024. 2, 7, 8, 17

    2024

  33. [41]

    Flux.1 kontext: Flow matching for in-context image generation and editing in latent space,

    Black Forest Labs, Stephen Batifol, Andreas Blattmann, Frederic Boesel, Saksham Consul, Cyril Diagne, Tim Dock- horn, Jack English, Zion English, Patrick Esser, Sumith Kulal, Kyle Lacey, Yam Levi, Cheng Li, Dominik Lorenz, Jonas M¨uller, Dustin Podell, Robin Rombach, Harry Sai...

  34. [42]

    Le, Tuan Pham, Sangho Lee, Christopher Clark, Aniruddha Kembhavi, Stephan Mandt, Ranjay Krishna, and Jiasen Lu

    Duong H. Le, Tuan Pham, Sangho Lee, Christopher Clark, Aniruddha Kembhavi, Stephan Mandt, Ranjay Krishna, and Jiasen Lu. One diffusion to generate them all. InProceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 2671–2682, 2025. 14

    2025

  35. [43]

    Scribble-guided diffusion for training- free text-to-image generation

    Seonho Lee, Jiho Choi, Seohyun Lim, Jiwook Kim, and Hyunjung Shim. Scribble-guided diffusion for training- free text-to-image generation. In2025 IEEE International Conference on Image Processing (ICIP), pages 1121–1126,

  36. [44]

    Control and realism: Best of both worlds in layout-to-image without training

    Bonan Li, Yinhan Hu, Songhua Liu, and Xinchao Wang. Control and realism: Best of both worlds in layout-to-image without training. InProceedings of International Conference on Machine Learning (ICML), 2025. 3

    2025

  37. [45]

    Blip-diffusion: Pre-trained subject representation for controllable text-to- image generation and editing

    Dongxu Li, Junnan Li, and Steven Hoi. Blip-diffusion: Pre-trained subject representation for controllable text-to- image generation and editing. InProceedings of Advances in Neural Information Processing Systems (NeurIPS), 2023. 14

    2023

  38. [46]

    Controlnet++: Improving conditional controls with efficient consistency feedback

    Ming Li, Taojiannan Yang, Huafeng Kuang, Jie Wu, Zhaon- ing Wang, Xuefeng Xiao, and Chen Chen. Controlnet++: Improving conditional controls with efficient consistency feedback. InProceedings of European Conference on Com- puter Vision (ECCV), 2024. 14

    2024

  39. [47]

    Gligen: Open-set grounded text-to-image generation

    Yuheng Li, Haotian Liu, Qingyang Wu, Fangzhou Mu, Jian- wei Yang, Jianfeng Gao, Chunyuan Li, and Yong Jae Lee. Gligen: Open-set grounded text-to-image generation. In Proceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2023. 14

    2023

  40. [48]

    Freecontrol: Efficient, training-free structural con- trol via one-step attention extraction

    Jiang Lin, Xinyu Chen, Song Wu, Zhiqiu Zhang, Jizhi Zhang, Ye Wang, Qiang Tang, qian Wang, Jian Yang, and Zili Yi. Freecontrol: Efficient, training-free structural con- trol via one-step attention extraction. InProceedings of Ad- vances in Neural Information Processing Systems...

  41. [49]

    Ctrl-x: Controlling structure and appear- ance for text-to-image generation without guidance

    Kuan Heng Lin, Sicheng Mo, Ben Klingher, Fangzhou Mu, and Bolei Zhou. Ctrl-x: Controlling structure and appear- ance for text-to-image generation without guidance. InPro- ceedings of Advances in Neural Information Processing Systems (NeurIPS), 2024. 2, 3, 4, 5, 6, 16, 17, 18, 20

    2024

  42. [50]

    Flow straight and fast: Learning to generate and transfer data with rectified flow, 2023

    Xingchao Liu, Chengyue Gong, and Qiang Liu. Flow straight and fast: Learning to generate and transfer data with rectified flow, 2023. 2

    2023

  43. [51]

    Dpm-solver: A fast ode solver for diffu- sion probabilistic model sampling in around 10 steps

    Cheng Lu, Yuhao Zhou, Fan Bao, Jianfei Chen, Chongxuan LI, and Jun Zhu. Dpm-solver: A fast ode solver for diffu- sion probabilistic model sampling in around 10 steps. In Proceedings of Advances in Neural Information Processing Systems (NeurIPS), 2022. 2

    2022

  44. [52]

    Latent consistency models: Synthesizing high- resolution images with few-step inference, 2023

    Simian Luo, Yiqin Tan, Longbo Huang, Jian Li, and Hang Zhao. Latent consistency models: Synthesizing high- resolution images with few-step inference, 2023. 18

    2023

  45. [53]

    Hpsv3: Towards wide-spectrum human pref- erence score

    Yuhang Ma, Yunhao Shui, Xiaoshi Wu, Keqiang Sun, and Hongsheng Li. Hpsv3: Towards wide-spectrum human pref- erence score. InProceedings of International Conference on Computer Vision (ICCV), 2025. 5

    2025

  46. [54]

    Sdedit: Guided image synthesis and editing with stochastic differential equa- tions

    Chenlin Meng, Yutong He, Yang Song, Jiaming Song, Jia- jun Wu, Jun-Yan Zhu, and Stefano Ermon. Sdedit: Guided image synthesis and editing with stochastic differential equa- tions. InProceedings of International Conference on Learn- ing Representations (ICLR), 2022. 3, 5, 6, 20

    2022

  47. [55]

    Freecontrol: Training-free spatial control of any text-to-image diffusion model with any condition

    Sicheng Mo, Fangzhou Mu, Kuan Heng Lin, Yanli Liu, Bochen Guan, Yin Li, and Bolei Zhou. Freecontrol: Training-free spatial control of any text-to-image diffusion model with any condition. InProceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2024. 2, 3...

    2024

  48. [56]

    T2i-adapter: Learn- ing adapters to dig out more controllable ability for text-to- image diffusion models

    Chong Mou, Xintao Wang, Liangbin Xie, Jian Zhang, Zhon- gang Qi, Ying Shan, and Xiaohu Qie. T2i-adapter: Learn- ing adapters to dig out more controllable ability for text-to- image diffusion models. InProceedings of AAAI Conference on Artificial Intelligence (AAAI), 2024. 5, 6...

    2024

  49. [57]

    K-lora: Unlock- ing training-free fusion of any subject and style loras

    Ziheng Ouyang, Zhen Li, and Qibin Hou. K-lora: Unlock- ing training-free fusion of any subject and style loras. In Proceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2025. 14

    2025

  50. [58]

    Semantic image synthesis with spatially-adaptive normalization

    Taesung Park, Ming-Yu Liu, Ting-Chun Wang, and Jun- Yan Zhu. Semantic image synthesis with spatially-adaptive normalization. InProceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2019. 14

    2019

  51. [59]

    Zero-shot image-to-image translation

    Gaurav Parmar, Krishna Kumar Singh, Richard Zhang, Yijun Li, Jingwan Lu, and Jun-Yan Zhu. Zero-shot image-to-image translation. InACM SIGGRAPH Conference on Computer Graphics and Interactive Techniques (SIGGRAPH), 2023. 3

    2023

  52. [60]

    Scalable diffusion models with transformers

    William Peebles and Saining Xie. Scalable diffusion models with transformers. InProceedings of the IEEE/CVF In- ternational Conference on Computer Vision (ICCV), pages 4195–4205, 2023. 2, 8

    2023

  53. [61]

    Pham, Jingye Chen, and Qifeng Chen

    Kien T. Pham, Jingye Chen, and Qifeng Chen. Tale: Training-free cross-domain image composition via adap- tive latent manipulation and energy-guided optimization. In Proceedings of the 32nd ACM International Conference on Multimedia, page 3160–3169, New York, NY , USA, 2024. As...

    2024

  54. [62]

    Orthogonal adaptation for modular customization of diffusion models

    Ryan Po, Guandao Yang, Kfir Aberman, and Gordon Wet- zstein. Orthogonal adaptation for modular customization of diffusion models. InProceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition (CVPR), pages 7964–7973, 2024. 14

    2024

  55. [63]

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

    Dustin Podell, Zion English, Kyle Lacey, Andreas Blattmann, Tim Dockhorn, Jonas M ¨uller, Joe Penna, and Robin Rombach. SDXL: Improving latent diffusion models for high-resolution image synthesis. InProceedings of Inter- national Conference on Learning Representations (ICLR),

  56. [64]

    Learn- ing transferable visual models from natural language su- pervision

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learn- ing transferable visual models from natural language su- pervision. InProceedings of International Conference on Ma...

    2021

  57. [65]

    Hierarchical text-conditional image genera- tion with clip latents.arXiv preprint arXiv:2204.06125, 1 (2):3, 2022

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

    2022 arXiv

  58. [66]

    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. InProceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2022. 2, 14

    2022

  59. [67]

    U-net: Convolutional networks for biomedical image segmentation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-net: Convolutional networks for biomedical image segmentation. InMedical Image Computing and Computer-Assisted In- tervention – MICCAI 2015, pages 234–241, Cham, 2015. Springer International Publishing. 2

    2015

  60. [68]

    Rb-modulation: Training-free stylization using reference-based modulation

    L Rout, Y Chen, N Ruiz, A Kumar, C Caramanis, S Shakkot- tai, and W Chu. Rb-modulation: Training-free stylization using reference-based modulation. InProceedings of Inter- national Conference on Learning Representations (ICLR),

  61. [69]

    Dreambooth: Fine tuning text-to-image diffusion models for subject-driven generation

    Nataniel Ruiz, Yuanzhen Li, Varun Jampani, Yael Pritch, Michael Rubinstein, and Kfir Aberman. Dreambooth: Fine tuning text-to-image diffusion models for subject-driven generation. InProceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2023. 14

    2023

  62. [70]

    Hyperdreambooth: Hypernetworks for fast personalization of text-to-image models

    Nataniel Ruiz, Yuanzhen Li, Varun Jampani, Wei Wei, Tingbo Hou, Yael Pritch, Neal Wadhwa, Michael Rubinstein, and Kfir Aberman. Hyperdreambooth: Hypernetworks for fast personalization of text-to-image models. InProceed- ings of the IEEE/CVF Conference on Computer Vision and Pa...

    2024

  63. [71]

    Palette: Image-to-image diffusion models

    Chitwan Saharia, William Chan, Huiwen Chang, Chris Lee, Jonathan Ho, Tim Salimans, David Fleet, and Mohammad Norouzi. Palette: Image-to-image diffusion models. In ACM SIGGRAPH Conference on Computer Graphics and Interactive Techniques (SIGGRAPH), 2022. 14

    2022

  64. [72]

    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. InProceedings of Advances in N...

    2022

  65. [73]

    Emu edit: Precise image editing via recognition and gener- ation tasks

    Shelly Sheynin, Adam Polyak, Uriel Singer, Yuval Kirstain, Amit Zohar, Oron Ashual, Devi Parikh, and Yaniv Taigman. Emu edit: Precise image editing via recognition and gener- ation tasks. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR...

    2024

  66. [74]

    De- noising diffusion implicit models

    Jiaming Song, Chenlin Meng, and Stefano Ermon. De- noising diffusion implicit models. InProceedings of Inter- national Conference on Learning Representations (ICLR),

  67. [75]

    Consistency models

    Yang Song, Prafulla Dhariwal, Mark Chen, and Ilya Sutskever. Consistency models. InProceedings of Inter- national Conference on Machine Learning (ICML), 2023. 2

    2023

  68. [76]

    Dual diffusion implicit bridges for image-to-image translation

    Xuan Su, Jiaming Song, Chenlin Meng, and Stefano Er- mon. Dual diffusion implicit bridges for image-to-image translation. InProceedings of International Conference on Learning Representations (ICLR), 2023. 3

    2023

  69. [77]

    Ominicontrol: Minimal and universal control for diffusion transformer

    Zhenxiong Tan, Songhua Liu, Xingyi Yang, Qiaochu Xue, and Xinchao Wang. Ominicontrol: Minimal and universal control for diffusion transformer. InProceedings of Interna- tional Conference on Computer Vision (ICCV), 2025. 14

    2025

  70. [78]

    Ominicontrol2: Efficient conditioning for diffusion transformers.arXiv preprint arXiv:2503.08280,

    Zhenxiong Tan, Qiaochu Xue, Xingyi Yang, Songhua Liu, and Xinchao Wang. Ominicontrol2: Efficient conditioning for diffusion transformers.arXiv preprint arXiv:2503.08280,

    arXiv

  71. [79]

    Add-it: Training-free object inser- tion in images with pretrained diffusion models

    Yoad Tewel, Rinon Gal, Dvir Samuel, Yuval Atzmon, Lior Wolf, and Gal Chechik. Add-it: Training-free object inser- tion in images with pretrained diffusion models. InProceed- ings of International Conference on Learning Representa- tions (ICLR), 2025. 3

    2025

  72. [80]

    Guide-and-rescale: Self- guidance mechanism for effective tuning-free real image editing

    Vadim Titov, Madina Khalmatova, Alexandra Ivanova, Dmitry Vetrov, and Aibek Alanov. Guide-and-rescale: Self- guidance mechanism for effective tuning-free real image editing. InProceedings of European Conference on Com- puter Vision (ECCV), 2024. 3

    2024

  73. [81]

    Splicing ViT features for semantic appearance transfer

    Narek Tumanyan, Omer Bar-Tal, Shai Bagon, and Tali Dekel. Splicing ViT features for semantic appearance transfer. In Proceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2022. 5, 14, 15, 18, 19

    2022

  74. [82]

    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 Conference on Com- puter Vision and Pattern Recognition (CVPR), pages 1921– 1930, 2023. 2, 3, 4, 5, 6, 14, 20

    1921

  75. [83]

    Diffusers: State-of-the-art diffu- sion models

    Patrick von Platen, Suraj Patil, Anton Lozhkov, Pedro Cuenca, Nathan Lambert, Kashif Rasul, Mishig Davaadorj, Dhruv Nair, Sayak Paul, William Berman, Yiyi Xu, Steven Liu, and Thomas Wolf. Diffusers: State-of-the-art diffu- sion models. https://github.com/huggingface/ diffusers...

    2022

  76. [84]

    Taming rectified flow for inversion and editing

    Jiangshan Wang, Junfu Pu, Zhongang Qi, Jiayi Guo, Yue Ma, Nisha Huang, Yuxin Chen, Xiu Li, and Ying Shan. Taming rectified flow for inversion and editing. InProceedings of International Conference on Machine Learning (ICML),

  77. [85]

    Fleet, Radu Soricut, Jason Baldridge, Mo- hammad Norouzi, Peter Anderson, and William Chan

    Su Wang, Chitwan Saharia, Ceslee Montgomery, Jordi Pont- Tuset, Shai Noy, Stefano Pellegrini, Yasumasa Onoe, Sarah Laszlo, David J. Fleet, Radu Soricut, Jason Baldridge, Mo- hammad Norouzi, Peter Anderson, and William Chan. Im- agen editor and editbench: Advancing and evaluati...

    2023

  78. [86]

    Instancediffusion: Instance- level control for image generation

    Xudong Wang, Trevor Darrell, Sai Saketh Rambhatla, Ro- hit Girdhar, and Ishan Misra. Instancediffusion: Instance- level control for image generation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 6232–6242, 2024. 14

    2024

  79. [87]

    Event-customized image generation

    Zhen Wang, Yilei JIANG, Dong Zheng, Jun Xiao, and Long Chen. Event-customized image generation. InProceedings of International Conference on Machine Learning (ICML),

  80. [88]

    Training-free dense-aligned diffusion guidance for modular conditional image synthesis

    Zixuan Wang, Duo Peng, Feng Chen, Yuwei Yang, and Yinjie Lei. Training-free dense-aligned diffusion guidance for modular conditional image synthesis. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 13135–13145, 2025. 3

    2025

  81. [89]

    Tianyi Wei, Yifan Zhou, Dongdong Chen, and Xingang Pan. Freeflux: Understanding and exploiting layer-specific roles in rope-based mmdit for versatile image editing.Pro- ceedings of International Conference on Computer Vision (ICCV), 2025. 3

    2025

  82. [90]

    Qwen-image technical report,

    Chenfei Wu, Jiahao Li, Jingren Zhou, Junyang Lin, Kaiyuan Gao, Kun Yan, Sheng ming Yin, Shuai Bai, Xiao Xu, Yilei Chen, Yuxiang Chen, Zecheng Tang, Zekai Zhang, Zhengyi Wang, An Yang, Bowen Yu, Chen Cheng, Dayiheng Liu, De- qing Li, Hang Zhang, Hao Meng, Hu Wei, Jingyuan Ni, K...

  83. [91]

    Omnigen2: Exploration to advanced multimodal generation,

    Chenyuan Wu, Pengfei Zheng, Ruiran Yan, Shitao Xiao, Xin Luo, Yueze Wang, Wanli Li, Xiyan Jiang, Yexin Liu, Junjie Zhou, Ze Liu, Ziyi Xia, Chaofan Li, Haoge Deng, Jiahao Wang, Kun Luo, Bo Zhang, Defu Lian, Xinlong Wang, Zhongyuan Wang, Tiejun Huang, and Zheng Liu. Omnigen2: Ex...

  84. [92]

    Human preference score v2: A solid benchmark for evaluating human preferences of text-to-image synthesis.arXiv preprint arXiv:2306.09341,

    Xiaoshi Wu, Yiming Hao, Keqiang Sun, Yixiong Chen, Feng Zhu, Rui Zhao, and Hongsheng Li. Human preference score v2: A solid benchmark for evaluating human preferences of text-to-image synthesis.arXiv preprint arXiv:2306.09341,

    arXiv

  85. [93]

    Dreamomni: Unified image generation and editing

    Bin Xia, Yuechen Zhang, Jingyao Li, Chengyao Wang, Yi- tong Wang, Xinglong Wu, Bei Yu, and Jiaya Jia. Dreamomni: Unified image generation and editing. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 28533–28543, 2025. 14

    2025

  86. [94]

    R&b: Region and boundary aware zero-shot grounded text-to-image generation

    Jiayu Xiao, Henglei Lv, Liang Li, Shuhui Wang, and Qing- ming Huang. R&b: Region and boundary aware zero-shot grounded text-to-image generation. InProceedings of Inter- national Conference on Learning Representations (ICLR),

  87. [95]

    Omnigen: Unified image gener- ation

    Shitao Xiao, Yueze Wang, Junjie Zhou, Huaying Yuan, Xin- grun Xing, Ruiran Yan, Chaofan Li, Shuting Wang, Tiejun Huang, and Zheng Liu. Omnigen: Unified image gener- ation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 13294–1...

    2025

  88. [96]

    Boxdiff: Text-to-image synthesis with training-free box-constrained diffusion

    Jinheng Xie, Yuexiang Li, Yawen Huang, Haozhe Liu, Wen- tian Zhang, Yefeng Zheng, and Mike Zheng Shou. Boxdiff: Text-to-image synthesis with training-free box-constrained diffusion. InProceedings of International Conference on Computer Vision (ICCV), 2023. 3

    2023

  89. [97]

    Anyrefill: A unified, data- efficient framework for left-prompt-guided vision tasks,

    Ming Xie, Chenjie Cao, Yunuo Cai, Xiangyang Xue, Yu- Gang Jiang, and Yanwei Fu. Anyrefill: A unified, data- efficient framework for left-prompt-guided vision tasks,

  90. [98]

    Imagere- ward: Learning and evaluating human preferences for text- to-image generation

    Jiazheng Xu, Xiao Liu, Yuchen Wu, Yuxuan Tong, Qinkai Li, Ming Ding, Jie Tang, and Yuxiao Dong. Imagere- ward: Learning and evaluating human preferences for text- to-image generation. InProceedings of Advances in Neural Information Processing Systems (NeurIPS), 2023. 5, 19

    2023

  91. [99]

    Unveil inversion and invariance in flow transformer for versatile image editing

    Pengcheng Xu, Boyuan Jiang, Xiaobin Hu, Donghao Luo, Qingdong He, Jiangning Zhang, Chengjie Wang, Yunsheng Wu, Charles Ling, and Boyu Wang. Unveil inversion and invariance in flow transformer for versatile image editing. In Proceedings of the IEEE/CVF Conference on Computer Vi...

  92. [100]

    Inversion-free image editing with natural language

    Sihan Xu, Yidong Huang, Jiayi Pan, Ziqiao Ma, and Joyce Chai. Inversion-free image editing with natural language. In Proceedings of Conference on Computer Vision and Pattern Recognition (CVPR), 2024. 3, 5, 6, 17, 18, 20

    2024

  93. [101]

    Restart sampling for improving generative processes

    Yilun Xu, Mingyang Deng, Xiang Cheng, Yonglong Tian, Ziming Liu, and Tommi Jaakkola. Restart sampling for improving generative processes. InProceedings of Advances in Neural Information Processing Systems (NeurIPS), 2023. 2, 5

    2023

  94. [102]

    Headrouter: A training-free image editing framework for mm-dits by adaptively routing attention heads, 2024

    Yu Xu, Fan Tang, Juan Cao, Yuxin Zhang, Xiaoyu Kong, Jintao Li, Oliver Deussen, and Tong-Yee Lee. Headrouter: A training-free image editing framework for mm-dits by adaptively routing attention heads, 2024. 3

    2024

  95. [103]

    CtrloRA: An extensible and efficient framework for con- trollable image generation

    Yifeng Xu, Zhenliang He, Shiguang Shan, and Xilin Chen. CtrloRA: An extensible and efficient framework for con- trollable image generation. InProceedings of International Conference on Learning Representations (ICLR), 2025. 14

    2025

  96. [104]

    Reco: Region-controlled text-to- image generation

    Zhengyuan Yang, Jianfeng Wang, Zhe Gan, Linjie Li, Kevin Lin, Chenfei Wu, Nan Duan, Zicheng Liu, Ce Liu, Michael Zeng, and Lijuan Wang. Reco: Region-controlled text-to- image generation. InProceedings of Conference on Com- puter Vision and Pattern Recognition (CVPR), pages 142...

    2023

  97. [105]

    Ip-adapter: Text compatible image prompt adapter for text- to-image diffusion models.arXiv preprint arxiv:2308.06721,

    Hu Ye, Jun Zhang, Sibo Liu, Xiao Han, and Wei Yang. Ip-adapter: Text compatible image prompt adapter for text- to-image diffusion models.arXiv preprint arxiv:2308.06721,

    arXiv

  98. [106]

    Adding conditional control to text-to-image diffusion models

    Lvmin Zhang, Anyi Rao, and Maneesh Agrawala. Adding conditional control to text-to-image diffusion models. In Proceedings of International Conference on Computer Vi- sion (ICCV), 2023. 2, 5, 6, 14, 20

    2023

  99. [107]

    The unreasonable effectiveness of deep features as a perceptual metric

    Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shechtman, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. InProceedings of Confer- ence on Computer Vision and Pattern Recognition (CVPR),

  100. [108]

    Ssr-encoder: Encoding selective subject representation for subject-driven generation

    Yuxuan Zhang, Yiren Song, Jiaming Liu, Rui Wang, Jinpeng Yu, Hao Tang, Huaxia Li, Xu Tang, Yao Hu, Han Pan, and Zhongliang Jing. Ssr-encoder: Encoding selective subject representation for subject-driven generation. InProceed- ings of the IEEE/CVF Conference on Computer Vision ...

    2024

  101. [109]

    Samcon- trol: Controlling pose and object for image editing with soft attention mask.ACM Trans

    Yue Zhang, Chao Wang, Feifei Fang, Yunzhi Zhuge, Hehe Fan, Xiaojun Chang, Cheng Deng, and Yi Yang. Samcon- trol: Controlling pose and object for image editing with soft attention mask.ACM Trans. Multimedia Comput. Commun. Appl., 2024. 3

    2024

  102. [110]

    Freecus: Free lunch subject-driven customization in diffu- sion transformers.Proceedings of International Conference on Computer Vision (ICCV), 2025

    Yanbing Zhang, Zhe Wang, Qin Zhou, and Mengping Yang. Freecus: Free lunch subject-driven customization in diffu- sion transformers.Proceedings of International Conference on Computer Vision (ICCV), 2025. 3

    2025

  103. [111]

    Easycontrol: Adding efficient and flexible control for diffusion transformer

    Yuxuan Zhang, Yirui Yuan, Yiren Song, Haofan Wang, and Jiaming Liu. Easycontrol: Adding efficient and flexible control for diffusion transformer. InProceedings of Interna- tional Conference on Computer Vision (ICCV), 2025. 14

    2025

  104. [112]

    Uni- controlnet: All-in-one control to text-to-image diffusion models

    Shihao Zhao, Dongdong Chen, Yen-Chun Chen, Jianmin Bao, Shaozhe Hao, Lu Yuan, and Kwan-Yee K Wong. Uni- controlnet: All-in-one control to text-to-image diffusion models. InProceedings of Advances in Neural Information Processing Systems (NeurIPS), 2023. 14

    2023

  105. [113]

    Unipc: A unified predictor-corrector framework for fast sampling of diffusion models

    Wenliang Zhao, Lujia Bai, Yongming Rao, Jie Zhou, and Jiwen Lu. Unipc: A unified predictor-corrector framework for fast sampling of diffusion models. InProceedings of Ad- vances in Neural Information Processing Systems (NeurIPS),

  106. [114]

    Local conditional controlling for text-to-image diffusion models.Proceedings of AAAI Conference on Artificial Intelligence (AAAI), 39(10), 2025

    Yibo Zhao, Liang Peng, Yang Yang, Zekai Luo, Hengjia Li, Yao Chen, Zheng Yang, Xiaofei He, Wei Zhao, Qinglin Lu, Wei Liu, and Boxi Wu. Local conditional controlling for text-to-image diffusion models.Proceedings of AAAI Conference on Artificial Intelligence (AAAI), 39(10), 2025. 14

    2025

  107. [115]

    Scene parsing through ade20k dataset

    Bolei Zhou, Hang Zhao, Xavier Puig, Sanja Fidler, Adela Barriuso, and Antonio Torralba. Scene parsing through ade20k dataset. InProceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017. 17

    2017

  108. [116]

    Replace objects in <d1> with semantically-related objects in ……

    Tianrui Zhu, Shiyi Zhang, Jiawei Shao, and Yansong Tang. Kv-edit: Training-free image editing for precise background preservation. InProceedings of International Conference on Computer Vision (ICCV), 2025. 3 We provide preliminaries in Sec. A and additional related work discus...

    2025

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.