Pith. sign in

REVIEW 3 major objections 8 minor 3 cited by

Enhancing Image Generation Fidelity via Progressive Prompts

T0 review · 3 major / 8 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read Splitting prompts into high- and low-level descriptions and feeding each image region its own masked cross-attention text state improves DiT image fidelity and prompt-following without retraining the base model.

desk verdict Ports masked regional cross-attention to DiT with an LLM prompt split, but the evaluation is unverifiable and the ablation contradicts the paper's own layer-depth story. read the letter →

arxiv 2501.07070 v1 pith:OYDSQ4NK submitted 2025-01-13 cs.CV

Zhen Xiong , Yuqi Li , Chuanguang Yang , Tiao Tan , Zhihong Zhu , Siyuan Li , Yue Ma This is my paper · ORCID
classification cs.CV
keywords Text-to-imagegenerationDiffusiontransformerRegionalpromptcontrolCross-attentionCoarse-to-fineLargelanguagemodelfollowingImagefidelity
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to show that diffusion transformer image generators can follow region-specific prompts without retraining the base model. Its pipeline, DiTPipe, uses a large language model to split a user's request into high-level and low-level prompts, then injects those prompts into different spatial regions through a masked cross-attention module. The authors report that this coarse-to-fine injection improves regional prompt-following and image fidelity compared with global-prompt baselines such as SDXL and SD-1.5. If the claim holds, any DiT-based text-to-image model could gain fine-grained regional control by swapping in a plug-in attention module.

What carries the argument

The load-bearing mechanism is the Controllable Region-Attention module, a masked regional variant of the standard cross-attention head. The image latent is divided into N adjacent regions; for each region a spatial mask is downsampled to latent resolution, flattened, and element-wise multiplied with the query. The masked query attends only to the text state of that region's positive prompt, producing a local feature, and all local features are summed. A global negative prompt is processed by a separate cross-attention branch and concatenated along the batch dimension, following the SDXL convention. The module replaces the native cross-attention in all 39 DiT blocks, and the progressive-prompt scheme feeds LLM-generated high-level and low-level descriptions through T5 and CLIP embeddings into this structure.

What would settle it

Take a fixed set of regional masks and prompts; run the pipeline twice, once injecting low-level prompts into shallow layers and once into deep layers. If the deep-layer version controls low-level details at least as well as the shallow-layer version, the coarse-to-fine layer-specialization premise is falsified, leaving the regional mask as the active ingredient.

Watch

Extended reading notes

Core claim

The paper claims that DiT image generators can be made region-controllable by decomposing a user prompt into high-level and low-level descriptions and giving each image region its own text state inside a masked cross-attention module. It reports that this coarse-to-fine injection, named DiTPipe, improves fidelity and prompt-following over global-prompt baselines, with higher PSNR/SSIM and lower LPIPS/rFID. It further claims that cross-attention layers at different depths play different roles, with deeper layers handling high-level content and shallow layers handling low-level details, which motivates where the progressive prompts are applied. The contribution is positioned as a plug-in replacement for the native cross-attention in Hunyuan-DiT, requiring no retraining of the base model.

Load-bearing premise

The pipeline's coarse-to-fine ordering depends on the claim that deep cross-attention layers control high-level content and shallow layers control low-level details, a premise the paper's own depth-ablation appears to contradict for low-level prompts.

Editorial extensions

If this is right

  • Replacing the native cross-attention with the masked module should give any DiT-based generator region-specific control without changing the base weights.
  • An LLM's high-level and low-level prompt split should be enough to drive coarse-to-fine generation, removing the need for per-region training or extra conditioning networks.
  • The scheme scales to 2, 4, or 9 regions by batching all positive prompt embeddings and summing the masked attention outputs.
  • Keeping the SDXL-style negative-prompt branch alongside regional positives should suppress unwanted global features while preserving local control.
  • The reported metrics (PSNR 28.2, SSIM 0.75, LPIPS 0.84, rFID 4.2) indicate gains in fidelity and distribution similarity over the global-prompt SDXL and SD-1.5 baselines.

Reading between the lines

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

  • The masking mechanism is agnostic to how regions are defined, so grid chunks could be replaced by semantic masks such as detected objects, segmentation maps, or user strokes without changing the module.
  • Because the modification touches only attention, the same regional-control idea should carry to other DiT-based text-to-image generators once their text-state packing is adapted.
  • A direct ablation of the paper's contribution would compare the masked regional attention fed with one merged prompt per region against the full high-level and low-level progressive split, isolating what the coarse-to-fine decomposition adds.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 8 minor

Summary. The paper proposes DiTPipe, a plug-in regional cross-attention mechanism for DiT-based text-to-image models, built on Hunyuan-DiT. The method uses an LLM to generate high-level and low-level prompts, divides the image into N regions, and injects each regional prompt through masked cross-attention in selected DiT blocks. The authors claim that deeper layers control high-level content and shallow layers control low-level detail, motivating a coarse-to-fine progressive prompt schedule. The evaluation consists of Table I (PSNR, SSIM, LPIPS, rFID against the vanilla Hunyuan-DiT and two SD-series models), qualitative comparisons in Figure 5, and an ablation in Figure 6.

Significance. If the central claims were adequately supported, the paper would offer a lightweight, training-free regional control mechanism for DiT image generators, which is a practically useful direction and complements existing UNet-based controllers such as GLIGEN. The availability of code and the focus on a DiT backbone are strengths. However, the empirical evidence does not currently establish the claimed benefit: the headline metrics are global and unverifiable, no metric measures regional prompt alignment, and the ablation in Figure 6 appears to contradict the layer-depth hypothesis that drives the method. The paper would require substantial additional experimentation and a reconciled narrative before its significance could be assessed.

major comments (3)
  1. [Section III-B, Table I] Table I and the surrounding text do not specify the evaluation protocol: no dataset or prompt set is named, no reference images are identified for PSNR/SSIM/LPIPS, no real-image distribution is identified for rFID, and no seeds or error bars are reported. More importantly, all five reported metrics are global image-quality or distributional scores; none of them measures whether the content of each region Ri follows its corresponding prompt Pi, which is the central claim of the paper. The comparison with the Couple method is described as quantitative in Section III-B, but no Couple numbers appear in Table I or anywhere else. Because the proposed method is specifically about regional controllability, the absence of a regional alignment metric (e.g., per-region CLIP similarity, detection-based object counts, or human evaluation of region-prompt correspondence) leaves the headline contribution unsupported by the quantitative results.
  2. [Section III-C, Figure 6] The ablation in Figure 6 contradicts the layer-depth premise stated in the Abstract and Section II. The paper claims that deeper cross-attention layers handle high-level content control while shallow layers handle low-level detail, and this claim motivates injecting high-level prompts into deeper layers and low-level prompts into shallow layers. However, the caption of Figure 6 reports that injecting low-level prompts into deeper layers gives better controllability. If that is the observed result, the coarse-to-fine progressive prompt assignment as described is not supported; if the caption is misstated, the experiment needs to be re-reported correctly. Either way, this is a load-bearing inconsistency because the progressive prompt design is one of the two core contributions.
  3. [Section III-C, implementation and ablations] The ablation described in the text says the authors gradually increased the number of Controllable Region-Attention modules from 0 to 39, but Figure 6 appears to show only injections at different depths with the total number of layers held constant. These are different experimental designs, and the paper does not report any quantitative result for either ablation (e.g., regional prompt-following accuracy as a function of the number of controlled blocks). Without such quantitative ablations, the claim that the module is responsible for the observed improvements is not established.
minor comments (8)
  1. [Table I] The column header 'PNSR' is a typo and should be 'PSNR'.
  2. [Abstract] The phrase 'in order for course-to-fine generation' should be 'coarse-to-fine generation'.
  3. [Section III-B] The word 'matrixes' should be 'metrics'.
  4. [Introduction] The text contains unresolved placeholder citations 'Some works [?], [?]' that should be replaced with actual references.
  5. [Section I] The sentence 'We perform extensive quantitative and qualitative results to proof the superiority' uses 'proof' where 'prove' is meant.
  6. [Section II-B] The phrase 'the required 333-length text states' is not explained; the reader is not told why the sequence length is 333 or how the CLIP embedding length is combined with the T5 embeddings.
  7. [Figure 5] The figure is hard to read: the rows are labeled 'Ours', 'XL', '1.5', but the correspondence between the displayed prompts and the regions is not clearly annotated, making it difficult to verify regional compliance from the figure alone.
  8. [Section III-A] The description says the model 'was trained using the SGM uniform scheduler', but no training data, training steps, or learning rate are described; if the model is not trained, this wording should be corrected.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the method is a plug-in architectural modification compared against external baselines, with no fitted parameter renamed as a prediction and no load-bearing self-citation chain.

full rationale

The paper's claimed contribution is an architectural plug-in: LLM-generated per-region prompts are injected through a masked regional cross-attention module in Hunyuan-DiT, and the resulting images are compared with global-prompt baselines (SDXL, SD-1.5, vanilla Hunyuan-DiT) in Table I and Figure 5. There is no fitted parameter whose value is later reported as a prediction, and no equation reduces the claimed result to its own input by construction. The coarse-to-fine layer-depth claim is presented as an empirical observation ('We discover that deeper layers always responsible for the high-level content control, while the shallow layers handles low-level content control') and is used to motivate the design; even if the ablation in Figure 6 appears to contradict that observation, that is an internal consistency or correctness issue, not circular reasoning. The paper cites several prior works by overlapping authors, but none of those citations supplies a uniqueness theorem or an unverified premise that forces the present conclusion; the central comparison is against external baselines. The absence of a per-region prompt-following metric and the lack of a specified reference protocol for PSNR/SSIM/LPIPS/rFID are serious evidence and reproducibility concerns, but they are not instances of self-definition, fitted-input-as-prediction, or renaming of a known result. The derivation chain is therefore self-contained rather than circular, so the appropriate score is 0.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

No fitted constants are reported; the method is an architectural modification. The load-bearing assumptions concern layer-depth semantics and mask fidelity, neither independently verified.

assumptions (3)
  • domain assumption Cross-attention layers at different depths have separable semantic roles: deep layers for high-level content, shallow layers for low-level detail.
    This motivates injecting progressive prompts at selected depths. The paper's Figure 6 ablation contradicts it: deeper injection of low-level prompts is reported as more controllable. Location: Abstract, Section II, Figure 6 caption.
  • domain assumption A mask down-sampled from image space to latent space preserves the intended region boundaries well enough for regional control.
    Section II-C uses downsampled masks to zero out queries outside each region; no analysis of mask boundary leakage is provided.
  • domain assumption Hunyuan-DiT's cross-attention accepts a concatenated T5 and CLIP text state of length 333, and this interface is unchanged by the proposed module.
    The method inherits the base model interface; any mismatch would invalidate the implementation. Location: Section II-B.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Enhancing Image Generation Fidelity via Progressive Prompts." pith.science (2026). https://pith.science/paper/OYDSQ4NK

@misc{pith2026250107070,
  author       = {Pith},
  title        = {Pith review of: Enhancing Image Generation Fidelity via Progressive Prompts},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OYDSQ4NK}},
  note         = {Machine review of arXiv:2501.07070}
}
read the original abstract

The diffusion transformer (DiT) architecture has attracted significant attention in image generation, achieving better fidelity, performance, and diversity. However, most existing DiT - based image generation methods focus on global - aware synthesis, and regional prompt control has been less explored. In this paper, we propose a coarse - to - fine generation pipeline for regional prompt - following generation. Specifically, we first utilize the powerful large language model (LLM) to generate both high - level descriptions of the image (such as content, topic, and objects) and low - level descriptions (such as details and style). Then, we explore the influence of cross - attention layers at different depths. We find that deeper layers are always responsible for high - level content control, while shallow layers handle low - level content control. Various prompts are injected into the proposed regional cross - attention control for coarse - to - fine generation. By using the proposed pipeline, we enhance the controllability of DiT - based image generation. Extensive quantitative and qualitative results show that our pipeline can improve the performance of the generated images.

Figures

Figures reproduced from arXiv: 2501.07070 by the authors.

Figure 1
Figure 1. Visual results of the proposed approach. We enable to generate the [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The details of proposed region-attention. We present the control [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 4
Figure 4. The process of prompt generation, we show the process of progressive [PITH_FULL_IMAGE:figures/full_fig_p002_4.png] view at source ↗
Figures from the paper (3 more)
Figure 3
Figure 3. Figure 3: The details of our proposed block. In the figure, we present the [PITH_FULL_IMAGE:figures/full_fig_p002_3.png]
Figure 5
Figure 5. Figure 5: Comparison results. We show the comparison with SDXL and SD-1.5. The left of figure is the low-level prompts. We perform the experiment on [PITH_FULL_IMAGE:figures/full_fig_p003_5.png]
Figure 6
Figure 6. Figure 6: The ablation study of our method applied to different depth of the model while keeping the total number of layers [PITH_FULL_IMAGE:figures/full_fig_p004_6.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Follow-Your-Creation: Empowering 4D Creation through Video Inpainting

    cs.CV 2025-06 conditional novelty 5.0 of 10

    Follow-Your-Creation fine-tunes the Wan2.1 video inpainting model on composite point-cloud and editing masks so a single monocular video can be converted into editable 4D video with new camera motion.

  2. Component Adaptive Clustering for Generalized Category Discovery

    cs.CV 2025-07 conditional novelty 4.0 of 10

    AdaGCD applies adaptive slot attention to decompose DINO image features into semantic components and pools them with global features, reporting SOTA accuracy on six GCD benchmarks.

  3. SkipVAR: Accelerating Visual Autoregressive Modeling via Adaptive Frequency-Aware Skipping

    cs.CV 2025-06 conditional novelty 4.0 of 10

    SkipVAR selects, per sample, between step skipping and unconditional branch replacement using handcrafted frequency features and a trained logistic regression, to accelerate visual autoregressive generation.

Reference graph

Works this paper leans on

38 extracted references · 7 canonical work pages · cited by 3 Pith papers

  1. [1]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” Advances in neural information processing systems , vol. 33, pp. 6840– 6851, 2020

    2020

  2. [2]

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

    D. Podell, Z. English, K. Lacey, A. Blattmann, T. Dockhorn, J. M ¨uller, J. Penna, and R. Rombach, “Sdxl: Improving latent diffusion models for high-resolution image synthesis,” arXiv preprint arXiv:2307.01952 , 2023

    arXiv 2023

  3. [3]

    High- resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High- resolution image synthesis with latent diffusion models,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 10 684–10 695

    2022

  4. [4]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark et al., “Learning transferable visual models from natural language supervision,” in International conference on machine learning . PMLR, 2021, pp. 8748–8763

    2021

  5. [5]

    Clip-kd: An empirical study of clip model distillation,

    C. Yang, Z. An, L. Huang, J. Bi, X. Yu, H. Yang, B. Diao, and Y . Xu, “Clip-kd: An empirical study of clip model distillation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 15 952–15 962

    2024

  6. [6]

    Relational diffusion distillation for efficient image generation,

    W. Feng, C. Yang, Z. An, L. Huang, B. Diao, F. Wang, and Y . Xu, “Relational diffusion distillation for efficient image generation,” in Proceedings of the 32nd ACM International Conference on Multimedia , 2024, pp. 205–213

    2024

  7. [7]

    Improving image generation with better captions,

    J. Betker, G. Goh, L. Jing, T. Brooks, J. Wang, L. Li, L. Ouyang, J. Zhuang, J. Lee, Y . Guoet al., “Improving image generation with better captions,” Computer Science. https://cdn. openai. com/papers/dall-e-3. pdf, vol. 2, no. 3, p. 8, 2023

    2023

  8. [8]

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

    C. Saharia, W. Chan, S. Saxena, L. Li, J. Whang, E. L. Denton, K. Ghasemipour, R. Gontijo Lopes, B. Karagol Ayan, T. Salimans et al., “Photorealistic text-to-image diffusion models with deep language understanding,” Advances in neural information processing systems , vol. 35, pp. 36 479–36 494, 2022

    2022

Show all 38 references
  1. [9]

    Hunyuan-dit: A powerful multi-resolution diffusion transformer with fine-grained chinese understanding,

    Z. Li, J. Zhang, Q. Lin, J. Xiong, Y . Long, X. Deng, Y . Zhang, X. Liu, M. Huang, Z. Xiao et al. , “Hunyuan-dit: A powerful multi-resolution diffusion transformer with fine-grained chinese understanding,” arXiv preprint arXiv:2405.08748, 2024

    2024 arXiv

  2. [10]

    Pixart-α: Fast training of diffusion transformer for photore- alistic text-to-image synthesis,

    J. Chen, J. Yu, C. Ge, L. Yao, E. Xie, Y . Wu, Z. Wang, J. Kwok, P. Luo, H. Lu et al., “Pixart-α: Fast training of diffusion transformer for photore- alistic text-to-image synthesis,” arXiv preprint arXiv:2310.00426, 2023

    2023 arXiv

  3. [11]

    Gra: Detecting oriented objects through group-wise rotating and attention,

    J. Wang, Y . Pu, Y . Han, J. Guo, Y . Wang, X. Li, and G. Huang, “Gra: Detecting oriented objects through group-wise rotating and attention,” arXiv preprint arXiv:2403.11127 , 2024

    2024 arXiv

  4. [12]

    Glide: a new approach for rapid, accurate docking and scoring. 2. enrichment factors in database screening,

    T. A. Halgren, R. B. Murphy, R. A. Friesner, H. S. Beard, L. L. Frye, W. T. Pollard, and J. L. Banks, “Glide: a new approach for rapid, accurate docking and scoring. 2. enrichment factors in database screening,” Journal of medicinal chemistry , vol. 47, no. 7, pp. 1750– 1759, 2004

    2004

  5. [13]

    Attention is all you need,

    A. Vaswani, “Attention is all you need,” Advances in Neural Information Processing Systems, 2017

    2017

  6. [14]

    Clipscore: A reference-free evaluation metric for image captioning,

    J. Hessel, A. Holtzman, M. Forbes, R. L. Bras, and Y . Choi, “Clipscore: A reference-free evaluation metric for image captioning,” arXiv preprint arXiv:2104.08718, 2021

    2021 arXiv

  7. [15]

    Adding conditional control to text-to-image diffusion models,

    L. Zhang, A. Rao, and M. Agrawala, “Adding conditional control to text-to-image diffusion models,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 3836–3847

    2023

  8. [16]

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

    Y . Li, H. Liu, Q. Wu, F. Mu, J. Yang, J. Gao, C. Li, and Y . J. Lee, “Gligen: Open-set grounded text-to-image generation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 22 511–22 521

    2023

  9. [17]

    Multidiffusion: Fusing diffusion paths for controlled image generation,

    O. Bar-Tal, L. Yariv, Y . Lipman, and T. Dekel, “Multidiffusion: Fusing diffusion paths for controlled image generation,” 2023

    2023

  10. [18]

    Taming rectified flow for inversion and editing,

    J. Wang, J. Pu, Z. Qi, J. Guo, Y . Ma, N. Huang, Y . Chen, X. Li, and Y . Shan, “Taming rectified flow for inversion and editing,”arXiv preprint arXiv:2411.04746, 2024

    2024 arXiv

  11. [19]

    Multibooth: Towards generating all your concepts in an image from text,

    C. Zhu, K. Li, Y . Ma, C. He, and L. Xiu, “Multibooth: Towards generating all your concepts in an image from text,” arXiv preprint arXiv:2404.14239, 2024

    2024 arXiv

  12. [20]

    Dino: Detr with improved denoising anchor boxes for end-to-end object detection,

    H. Zhang, F. Li, S. Liu, L. Zhang, H. Su, J. Zhu, L. M. Ni, and H.-Y . Shum, “Dino: Detr with improved denoising anchor boxes for end-to-end object detection,” arXiv preprint arXiv:2203.03605 , 2022

    2022 arXiv

  13. [21]

    Cross- image relational knowledge distillation for semantic segmentation,

    C. Yang, H. Zhou, Z. An, X. Jiang, Y . Xu, and Q. Zhang, “Cross- image relational knowledge distillation for semantic segmentation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 12 319–12 328

    2022

  14. [22]

    Sglp: A similarity guided fast layer partition pruning for compressing large deep models,

    Y . Li, Y . Lu, Z. Dong, C. Yang, Y . Chen, and J. Gou, “Sglp: A similarity guided fast layer partition pruning for compressing large deep models,” arXiv preprint arXiv:2410.14720 , 2024

    2024

  15. [23]

    Comae: Comprehensive attribute exploration for zero-shot hashing,

    Y . Li, Q. Long, Y . Zhou, N. Cao, S. Liu, F. Zheng, Z. Zhu, Z. Ning, M. Xiao, X. Wang et al., “Comae: Comprehensive attribute exploration for zero-shot hashing,” arXiv preprint arXiv:2402.16424 , 2024

    2024 arXiv

  16. [24]

    Follow your pose: Pose-guided text-to-video generation using pose-free videos,

    Y . Ma, Y . He, X. Cun, X. Wang, S. Chen, X. Li, and Q. Chen, “Follow your pose: Pose-guided text-to-video generation using pose-free videos,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 38, no. 5, 2024, pp. 4117–4125

    2024

  17. [25]

    Follow-your-click: Open-domain regional image animation via short prompts,

    Y . Ma, Y . He, H. Wang, A. Wang, C. Qi, C. Cai, X. Li, Z. Li, H.-Y . Shum, W. Liu et al. , “Follow-your-click: Open-domain regional image animation via short prompts,” arXiv preprint arXiv:2403.08268 , 2024

    2024 arXiv

  18. [26]

    Visual knowledge graph for human action reasoning in videos,

    Y . Ma, Y . Wang, Y . Wu, Z. Lyu, S. Chen, X. Li, and Y . Qiao, “Visual knowledge graph for human action reasoning in videos,” in Proceedings of the 30th ACM International Conference on Multimedia , 2022, pp. 4132–4141

    2022

  19. [27]

    Magicstick: Controllable video editing via control handle transforma- tions,

    Y . Ma, X. Cun, Y . He, C. Qi, X. Wang, Y . Shan, X. Li, and Q. Chen, “Magicstick: Controllable video editing via control handle transforma- tions,” arXiv preprint arXiv:2312.03047 , 2023

    2023 arXiv

  20. [28]

    Follow-your-emoji: Fine-controllable and ex- pressive freestyle portrait animation,

    Y . Ma, H. Liu, H. Wang, H. Pan, Y . He, J. Yuan, A. Zeng, C. Cai, H.- Y . Shum, W. Liu et al. , “Follow-your-emoji: Fine-controllable and ex- pressive freestyle portrait animation,” arXiv preprint arXiv:2406.01900 , 2024

    2024 arXiv

  21. [29]

    Follow-your-canvas: Higher-resolution video outpainting with extensive content generation,

    Q. Chen, Y . Ma, H. Wang, J. Yuan, W. Zhao, Q. Tian, H. Wang, S. Min, Q. Chen, and W. Liu, “Follow-your-canvas: Higher-resolution video outpainting with extensive content generation,” arXiv preprint arXiv:2409.01055, 2024

    2024 arXiv

  22. [30]

    Cove: Unleashing the diffusion feature correspondence for consistent video editing,

    J. Wang, Y . Ma, J. Guo, Y . Xiao, G. Huang, and X. Li, “Cove: Unleashing the diffusion feature correspondence for consistent video editing,” arXiv preprint arXiv:2406.08850 , 2024

    2024 arXiv

  23. [31]

    Instantswap: Fast customized concept swapping across sharp shape differences,

    C. Zhu, K. Li, Y . Ma, L. Tang, C. Fang, C. Chen, Q. Chen, and X. Li, “Instantswap: Fast customized concept swapping across sharp shape differences,” arXiv preprint arXiv:2412.01197 , 2024

    2024 arXiv

  24. [32]

    Dit4edit: Diffusion transformer for image editing,

    K. Feng, Y . Ma, B. Wang, C. Qi, H. Chen, Q. Chen, and Z. Wang, “Dit4edit: Diffusion transformer for image editing,” arXiv preprint arXiv:2411.03286, 2024

    2024 arXiv

  25. [33]

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

    C. Raffel, N. Shazeer, A. Roberts, K. Lee, S. Narang, M. Matena, Y . Zhou, W. Li, and P. J. Liu, “Exploring the limits of transfer learning with a unified text-to-text transformer,” Journal of machine learning research, vol. 21, no. 140, pp. 1–67, 2020

    2020

  26. [34]

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

    N. Ruiz, Y . Li, V . Jampani, Y . Pritch, M. Rubinstein, and K. Aberman, “Dreambooth: Fine tuning text-to-image diffusion models for subject- driven generation,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2023, pp. 22 500–22 510

    2023

  27. [35]

    Ip-adapter: Text compatible image prompt adapter for text-to-image diffusion models,

    H. Ye, J. Zhang, S. Liu, X. Han, and W. Yang, “Ip-adapter: Text compatible image prompt adapter for text-to-image diffusion models,” arXiv preprint arXiv:2308.06721 , 2023

    2023 arXiv

  28. [36]

    Discrepancy and structure-based contrast for test-time adaptive retrieval,

    Z. Ma, Y . Li, Y . Luo, X. Luo, J. Li, C. Chen, X.-S. Hua, and G. Lu, “Discrepancy and structure-based contrast for test-time adaptive retrieval,” IEEE Transactions on Multimedia , 2024

    2024

  29. [37]

    A generic layer pruning method for signal modulation recognition deep learning models,

    Y . Lu, Y . Zhu, Y . Li, D. Xu, Y . Lin, Q. Xuan, and X. Yang, “A generic layer pruning method for signal modulation recognition deep learning models,” arXiv preprint arXiv:2406.07929 , 2024

    2024 arXiv

  30. [38]

    Scalable diffusion models with transformers,

    W. Peebles and S. Xie, “Scalable diffusion models with transformers,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 4195–4205

    2023

Pith tools

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