Pith. sign in

REVIEW 3 major objections 5 minor 72 references

Stable-Hair v2: Real-World Hair Transfer via Multiple-View Diffusion Model

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 Diffusion model transfers hairstyles across multiple views.

desk verdict Genuinely new multi-view hair transfer framework, but the paper's own Table II contradicts its headline claim of superiority. read the letter →

arxiv 2507.07591 v1 pith:YWC3ROYS submitted 2025-07-10 cs.CV

Kuiyuan Sun , Yuxuan Zhang , Jichao Zhang , Jiaming Liu , Wei Wang , Niculae Sebe , Yao Zhao This is my paper
classification cs.CV
keywords multi-viewhairtransferdiffusionmodelsavatargenerationhairstyleviewconsistencyposeconditioningtemporalattentionsynthetictrainingdata
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper tries to establish that multi-view hair transfer—placing a reference hairstyle onto a source face while preserving identity and background from every angle—can be done with a diffusion model trained on synthetic multi-view data. It argues this is the first such system and reports that it outperforms existing single-view hair transfer methods that are piped through a multi-view generator. The practical payoff, if true, is that digital humans and avatars could get consistent hair styling from one reference photo instead of per-view editing. The paper also presents a synthetic data pipeline that produces bald images, reference hairstyles, and view-aligned source–bald pairs for training and evaluation.

What carries the argument

The load-bearing machinery is the combination of a synthetic multi-view training pipeline and a staged training strategy. The face-finetuned multi-view diffusion model (SV3D fine-tuned on 3D GAN facial videos) supplies view-aligned source–bald pairs, and the polar–azimuth pose embedding inside a Latent IdentityNet supplies viewpoint control. A trainable Hair Extractor, whose self-attention features become keys and values for new cross-attention layers, supplies hairstyle content, and temporal self-attention blocks added to the U-Net supply inter-frame consistency. Training is staged so pose control is learned on bald images first, then hair transfer, then temporal smoothing.

What would settle it

Render the fine-tuned multi-view model's 21 outputs for real FFHQ images and compare them against ground-truth multi-view portrait captures of the same people, measuring head-pose error and identity similarity; if the synthetic pairs show systematic pose or identity drift, the training-data foundation of the method collapses. A simpler check is to run the full pipeline on images with extreme or unusual viewpoints and count how often the multi-view output loses identity or produces inconsistent head shapes.

Watch

Extended reading notes

Core claim

The paper's central claim is that a latent diffusion model can generate photorealistic, view-consistent hair transfer by learning in three stages: first a pose-controllable Latent IdentityNet learns to rotate bald heads using polar-azimuth pose embeddings; then a Hair Extractor injects reference hair features through cross-attention layers; and finally temporal attention layers smooth transitions between consecutive frames. Training data comes from a face-finetuned multi-view diffusion model that produces 21 views for each source and bald image, where the fine-tuning itself uses roughly 20,000 multi-view facial videos synthesized by a 3D-aware GAN. The paper reports quantitative gains over existing methods on single-view and multi-view hair transfer, and user-study wins on transfer accuracy, preservation, and naturalness.

Load-bearing premise

The whole system rests on the assumption that synthetic multi-view face videos generated by a 3D GAN, and then used to fine-tune a multi-view diffusion model, produce pose-aligned and identity-preserving pairs accurate enough to train and evaluate a real-world hair transfer model, with no validation against real multi-view portrait captures.

Editorial extensions

If this is right

  • Hair styling for avatars and digital humans could be generated from a single reference photo with consistent results across viewpoints.
  • The staged recipe—pose control on bald heads first, then hair, then temporal smoothing—can be reused for other attribute-transfer tasks where pose and content compete.
  • The synthetic data pipeline removes the need for a large real multi-view hair dataset, making multi-view transfer practical where such data is unavailable.
  • End-to-end multi-view transfer improves consistency metrics over the baseline of single-view transfer followed by multi-view generation, establishing a benchmark for this task.
  • The temporal attention layer keeps head motion close to ground truth, a step toward video-level hair transfer.

Reading between the lines

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

  • If the synthetic-to-real gap is small, the same data pipeline could extend from front-facing views to full 360-degree views by adding more training viewpoints from the 3D GAN.
  • The staged training strategy may matter more than the specific hair domain; it could transfer to multi-view makeup, beard, or accessory editing where pose and attribute content compete.
  • The reported background-reconstruction weakness suggests a testable extension: explicitly conditioning on background features should reduce inter-frame flicker, measurable with frame-difference heatmaps.
  • One can probe whether the Hair Extractor truly learns hairstyle rather than identity by feeding reference images with identical hairstyles on very different faces and checking whether transfer remains invariant.
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 Stable-Hair v2, a multi-view hair transfer framework built on a latent diffusion backbone. The main contributions are: a training-data generation pipeline that produces triplets of source, reference, and bald images together with multi-view pairs, using a diffusion-based bald converter, a data-augmented inpainting model, and a face-finetuned SV3D model; a pose-controllable latent IdentityNet for view control; a hair extractor with cross-attention; temporal attention layers for view consistency; and a three-stage training strategy. The authors claim this is the first work to apply multi-view diffusion models to hair transfer and that it significantly outperforms existing methods, based on single-view and multi-view quantitative comparisons, qualitative visualizations, user studies, and ablations.

Significance. The single-view results, the detailed data-generation pipeline, and the modular three-stage training strategy are useful contributions, and the authors make the code publicly available. If the multi-view claims held, this would be a meaningful step toward practical avatar and digital-human applications. However, the central multi-view superiority claim is directly contradicted by the paper's own Table II, and the quantitative evaluation is performed on a dataset generated by the same synthetic pipeline used for training, which substantially weakens the 'real-world' and state-of-the-art assertions as currently written.

major comments (3)
  1. [Section IV.C.2, Table II] The text states that 'Our method not only addresses the limitations of existing approaches but also demonstrates superior performance in terms of FID and IDS.' This is not supported by Table II: with FID (lower is better), Stable-Hair v2 scores 32.170, worse than HairFusion+SV3D (31.203) and HairCLIPV2+SV3D (29.652). The method is also worse on PSNR (26.347 vs. 27.846 for HairFusion+SV3D) and SSIM (0.490 vs. 0.564). Only CLIP-I (0.411 vs. 0.406) and IDS (0.683 vs. 0.659) show small advantages over the stronger baseline. The abstract's claim of 'significantly outperforming existing methods' is therefore internally inconsistent with the reported multi-view numbers. The authors must either correct the table or the prose, or rerun the multi-view evaluation; as presented, the multi-view SOTA claim is false on three of the five headline metrics.
  2. [Section IV.B and Section IV.C.2] The quantitative multi-view evaluation uses 'a dataset generated by our training data pipeline' (2,500 multi-view sets). Because the proposed model is trained on the same synthetic SV3D/GRAM-HD distribution used to create this test set, while the comparison pipelines (HairCLIPV2/SV3D, HairFusion/SV3D) are applied post hoc to real images, the comparison is biased in favor of the proposed method. The paper does not report whether test identities were held out from the approximately 20,000 training sets, nor does it validate the synthetic multi-view data against real multi-view portrait captures. The authors should evaluate on real multi-view data or at minimum provide disjoint train/test splits and an analysis of the synthetic-to-real gap.
  3. [Section III.B.3 and Section IV.A] The entire training and evaluation pipeline depends on the face-finetuned SV3D model, which is fine-tuned on roughly 20,000 GRAM-HD-generated multi-view face videos and then used to produce 21 views for both source and bald images. The manuscript offers no validation that these generated multi-view pairs are view-aligned, identity-preserving, or geometrically accurate. Since the method is claimed to work in 'real-world' settings, the authors should provide evidence that the synthetic multi-view data is a faithful proxy for real multi-view captures, for example by comparing generated views with real multi-view portrait data or by reporting pose/identity accuracy metrics on held-out real images.
minor comments (5)
  1. [Equation (1)] The LDM loss in Eq. (1) is written as E(||epsilon - epsilon_theta(zt,t,c)||), missing the squared L2 norm and the subscript 2 used in the standard formulation; please align with the original LDM objective.
  2. [Section III.C.1 and Fig. 9 / Table III] The naming is inconsistent: the module is called 'Pose-Controllable Latent IdentityNet' in the text but 'Latent ControlNet' and 'pixel' vs. 'latent' in Fig. 9 and Table III. Please unify the terminology.
  3. [Equation (2)] The symbol epsilon is used both for the noise in Eq. (1) and for the 'noise augmentation of camera pose' in Eq. (2); this is confusing and should be clarified with distinct notation.
  4. [Section IV.A vs. Section IV.B] The test set is described as '50 distinct samples' in Section IV.B but the quantitative experiments in Section IV.C.2 mention 2,500 sets of multi-view images; the relationship between these numbers should be clarified.
  5. [References [15] and [19]] References [15] and [19] appear to be the same paper (Chung et al., AAAI 2025) with different titles and venues; please deduplicate and provide the correct citation.

Circularity Check

1 steps flagged · score 5.0 of 10

The multi-view quantitative benchmark is generated by the same synthetic pipeline used to train the model, so the 'real-world' multi-view claim rests on in-distribution reproduction of the authors' own data.

  1. fitted input called prediction [Section IV-C.2 (Quantitative Comparison), Table II; Section IV-A Implementation Details; Section III-B.3]
    "For the quantitative experiments, we utilized a dataset generated by our training data pipeline."

    The evaluation set for Table II is produced by the same pipeline (diffusion-based bald converter plus face-finetuned SV3D multi-view generator) that produced the 'approximately 20,000 high-quality multi-view image pairs' used for training. The model's Stage-2 training objective uses exactly such pipeline-generated frames as ground truth, and the quantitative benchmark reuses the same distribution as that ground truth. Consequently the multi-view 'predictions' are compared against the outputs of the very generator the model was fitted to imitate; no independent real multi-view capture validates the 'Real-World Hair Transfer' claim. This is a benchmark that reduces to the training target rather than an external test.

full rationale

No derivation-equation circularity was found: Eq. (2) and Eq. (3) define pose conditioning and the noise-prediction loss without smuggling the target into the inputs, and the single-view experiments use real FFHQ images with standard metrics against external baselines. The main circularity concern is the multi-view evaluation, which uses 'a dataset generated by our training data pipeline' (Section IV-C.2). Since the same face-finetuned SV3D generator supplies both the training triplets and the test-set ground truth, the quantitative multi-view results (Table II, Fig. 7) measure how well the model reproduces the authors' own synthetic pipeline, not real-world multi-view consistency. In addition, Table II's numbers do not support the abstract's claim of 'significantly outperforming existing methods' on the multi-view task: the method is worse than HairFusion+SV3D on FID (32.170 vs 31.203), PSNR (26.347 vs 27.846), and SSIM (0.490 vs 0.564); this is a correctness and reporting inconsistency, not a circularity issue. Because the single-view contribution has independent empirical content, the score is moderate rather than high.

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

Stable-Hair v2 introduces no physical or model entity beyond learned neural modules (Latent IdentityNet, Hair Extractor, Temporal Attention). The free parameters listed are hand-chosen training and inference hyperparameters. The strongest assumptions are about the synthetic multi-view data pipeline: that face-finetuned SV3D on GAN-generated videos yields reliable multi-view ground truth for real faces, and that evaluating on data generated by the same pipeline is a valid benchmark.

free parameters (3)
  • classifier-free guidance scale = 1.5
    Inference hyperparameter chosen in Section IV-A with no reported sensitivity analysis; directly affects fidelity and view consistency.
  • temporal sequence length k = 12
    Number of consecutive frames sampled during temporal attention training (Section III-E); no ablation of this length.
  • number of views K = 21
    The fine-tuned SV3D model generates 21 viewpoints for each source and bald image (Section IV-A); the impact of K on consistency is not studied.
assumptions (3)
  • domain assumption Face-finetuned SV3D, trained on about 20,000 synthetic GRAM-HD multi-view videos, produces view-aligned and identity-preserving multi-view pairs for real FFHQ source and bald images.
    Used in Section III-B.3 to create all training and evaluation triplets; the accuracy of these synthetic views is not validated against real multi-view captures.
  • domain assumption Synthetic multi-view images generated by the authors' own pipeline can serve as ground truth for both training and evaluation.
    Section IV.B states 'we utilized a dataset generated by our training data pipeline', so the benchmark distribution is the same as the training distribution.
  • domain assumption Adding sinusoidal polar and azimuth embeddings to the time embedding is a sufficient camera-pose conditioning mechanism.
    Equation (2) in Section III-C.1 introduces this pose conditioning; pose control may be conflated with head-shape variations learned from SV3D outputs.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Stable-Hair v2: Real-World Hair Transfer via Multiple-View Diffusion Model." pith.science (2026). https://pith.science/paper/YWC3ROYS

@misc{pith2026250707591,
  author       = {Pith},
  title        = {Pith review of: Stable-Hair v2: Real-World Hair Transfer via Multiple-View Diffusion Model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YWC3ROYS}},
  note         = {Machine review of arXiv:2507.07591}
}
read the original abstract

While diffusion-based methods have shown impressive capabilities in capturing diverse and complex hairstyles, their ability to generate consistent and high-quality multi-view outputs -- crucial for real-world applications such as digital humans and virtual avatars -- remains underexplored. In this paper, we propose Stable-Hair v2, a novel diffusion-based multi-view hair transfer framework. To the best of our knowledge, this is the first work to leverage multi-view diffusion models for robust, high-fidelity, and view-consistent hair transfer across multiple perspectives. We introduce a comprehensive multi-view training data generation pipeline comprising a diffusion-based Bald Converter, a data-augment inpainting model, and a face-finetuned multi-view diffusion model to generate high-quality triplet data, including bald images, reference hairstyles, and view-aligned source-bald pairs. Our multi-view hair transfer model integrates polar-azimuth embeddings for pose conditioning and temporal attention layers to ensure smooth transitions between views. To optimize this model, we design a novel multi-stage training strategy consisting of pose-controllable latent IdentityNet training, hair extractor training, and temporal attention training. Extensive experiments demonstrate that our method accurately transfers detailed and realistic hairstyles to source subjects while achieving seamless and consistent results across views, significantly outperforming existing methods and establishing a new benchmark in multi-view hair transfer. Code is publicly available at https://github.com/sunkymepro/StableHairV2.

Figures

Figures reproduced from arXiv: 2507.07591 by the authors.

Figure 1
Figure 1. Given a source and a reference image, Stable-Hair v2 precisely transfers key hair attributes—color, length, curvature, and overall structure—onto the [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The proposed multi-view hair transfer model employs a multiple stage training strategy: (1) training a Pose-Controllable Latent IdentityNet with [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. The reference images generated with and without data augmentation [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Multi-View Training Data Generation Pipeline: This pipeline aims to generate {Multiple-View Original Image, Reference Image, Multiple-View Bald Image} triplets for training. The pipeline includes three key steps. Step 1: use Diffusion-based Bald Convertor to convert or…
Figure 5
Figure 5. Figure 5: Qualitative comparison of different methods. Compared to other approaches, our method achieves more refined and stable hairstyle transfer without [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Qualitative comparison of different multi-view hair transfer pipeline. Compared to other pipeline, our method achieves more refined and stable hairstyle [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: The head motion along the x-axis and y-axis of the three-dimensional coordinate system of three methods and ground truth. [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Visualization of mult-view hair transfer. Our methods can not only transfer ref hairstyle to the original img, but also generate corresponding multi-view [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 9
Figure 9. Figure 9: Visual results of the ablation study comparing different architectural [PITH_FULL_IMAGE:figures/full_fig_p011_9.png]
Figure 11
Figure 11. Figure 11: Ablation comparison between joint and two-stage training strategies. [PITH_FULL_IMAGE:figures/full_fig_p011_11.png]
Figure 12
Figure 12. Figure 12: Visual Ablation results of temporal attention. The first four columns represent consecutive frames from the video, while the fourth column displays the heatmap of frame-wise changes. The heatmap for the video without temporal attention shows regions with higher bright…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

72 extracted references · 50 canonical work pages

  1. [1]

    Michigan: Multi-input-conditioned hair image generation for portrait editing,

    Z. Tan, M. Chai, D. Chen, J. Liao, Q. Chu, L. Yuan, S. Tulyakov, and N. Yu, “Michigan: Multi-input-conditioned hair image generation for portrait editing,” ACM Transactions on Graphics (TOG) , vol. 39, no. 4, pp. 1–13, 2020

    work page 2020

  2. [2]

    Hair-GANs: Recovering 3D Hair Structure from a Single Image,

    M. Zhang and Y . Zheng, “Hair-GANs: Recovering 3D Hair Structure from a Single Image,” arXiv e-prints, p. arXiv:1811.06229, Nov. 2018

    arXiv 2018

  3. [3]

    Gan with multivariate disentangling for controllable hair editing,

    X. Guo, M. Kan, T. Chen, and S. Shan, “Gan with multivariate disentangling for controllable hair editing,” ECCV, 2022

    work page 2022

  4. [4]

    Hairnet: Hairstyle transfer withpose changes,

    P. Zhu, R. Abdal, J. Femiani, and P. Wonka, “Hairnet: Hairstyle transfer withpose changes,” ECCV, 2022

    work page 2022

  5. [5]

    Hairnerf: Geometry-aware image synthesis for hairstyle transfer,

    S. Chang, G. Kim, and H. Kim, “Hairnerf: Geometry-aware image synthesis for hairstyle transfer,” in Proceedings of the IEEE/CVF In- ternational Conference on Computer Vision , 2023, pp. 2448–2458

    work page 2023

  6. [6]

    Hairfit: pose-invariant hairstyle transfer via flow-based hair alignment and semantic-region-aware inpainting,

    C. Chung, T. Kim, H. Nam, S. Choi, G. Gu, S. Park, and J. Choo, “Hairfit: pose-invariant hairstyle transfer via flow-based hair alignment and semantic-region-aware inpainting,” BMVC, 2021

    work page 2021

  7. [7]

    Barbershop: Gan-based image compositing using segmentation masks,

    P. Zhu, R. Abdal, J. Femiani, and P. Wonka, “Barbershop: Gan-based image compositing using segmentation masks,” 2021

    work page 2021

  8. [8]

    Loho: Latent optimization of hairstyles via orthogonaliza- tion,

    Saha, Rohit, Duke, Brendan, Shkurti, Florian, Taylor, Graham, Aarabi, and Parham, “Loho: Latent optimization of hairstyles via orthogonaliza- tion,” in CVPR, 2021

    work page 2021

Show all 72 references
  1. [9]

    Hairclip: Design your hair by text and reference image,

    T. Wei, D. Chen, W. Zhou, J. Liao, Z. Tan, L. Yuan, W. Zhang, and N. Yu, “Hairclip: Design your hair by text and reference image,” Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022

    2022

  2. [10]

    Hairclipv2: Unifying hair editing via proxy feature blending,

    T. Wei, D. Chen, W. Zhou, J. Liao, W. Zhang, G. Hua, and N. Yu, “Hairclipv2: Unifying hair editing via proxy feature blending,” ICCV, 2023

    2023

  3. [11]

    Hairfastgan: Realistic and robust hair transfer with a fast encoder-based approach,

    M. Nikolaev, M. Kuznetsov, D. Vetrov, and A. Alanov, “Hairfastgan: Realistic and robust hair transfer with a fast encoder-based approach,” NeurIPS, 2024

    2024

  4. [12]

    Stylegan salon: Multi-view latent optimization for pose-invariant hairstyle transfer,

    S. Khwanmuang, P. Phongthawee, P. Sangkloy, and S. Suwajanakorn, “Stylegan salon: Multi-view latent optimization for pose-invariant hairstyle transfer,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2023

    2023

  5. [13]

    Style your hair: Latent optimization for pose-invariant hairstyle transfer via local- style-aware hair alignment,

    T. Kim, C. Chung, Y . Kim, S. Park, K. Kim, and J. Choo, “Style your hair: Latent optimization for pose-invariant hairstyle transfer via local- style-aware hair alignment,” ECCV, 2022

    2022

  6. [14]

    Stable- hair: Real-world hair transfer via diffusion model,

    Y . Zhang, Q. Zhang, Y . Song, J. Zhang, H. Tang, and J. Liu, “Stable- hair: Real-world hair transfer via diffusion model,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 39, no. 10, 2025, pp. 10 348–10 356

    2025

  7. [15]

    What to preserve and what to transfer: Faithful, identity-preserving diffusion-based hairstyle transfer,

    C. Chung, S. Park, J. Kim, and J. Choo, “What to preserve and what to transfer: Faithful, identity-preserving diffusion-based hairstyle transfer,” in The Association for the Advancement of Artificial Intelligence , 2025

    2025

  8. [16]

    Hairmapper: Removing hair from portraits using gans,

    Y . Wu, Y .-L. Yang, and X. Jin, “Hairmapper: Removing hair from portraits using gans,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2022, pp. 4227– 4236

    2022

  9. [17]

    SV3D: Novel multi-view synthesis and 3D generation from a single image using latent video diffusion,

    V . V oleti, C.-H. Yao, M. Boss, A. Letts, D. Pankratz, D. Tochilkin, C. Laforte, R. Rombach, and V . Jampani, “SV3D: Novel multi-view synthesis and 3D generation from a single image using latent video diffusion,” in European Conference on Computer Vision (ECCV) , 2024. 13

    2024

  10. [18]

    Few-shot head swapping in the wild,

    C. Shu, H. Wu, H. Zhou, J. Liu, Z. Hong, C. Ding, J. Han, J. Liu, E. Ding, and J. Wang, “Few-shot head swapping in the wild,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 10 789–10 798

    2022

  11. [19]

    What to preserve and what to transfer: Faithful, identity-preserving diffusion-based hairstyle transfer,

    C. Chung, S. Park, J. Kim, and J. Choo, “What to preserve and what to transfer: Faithful, identity-preserving diffusion-based hairstyle transfer,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 39, no. 3, 2025, pp. 2582–2590

    2025

  12. [20]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” arXiv preprint arxiv:2006.11239 , 2020

    2006 arXiv

  13. [21]

    Hierarchical text-conditional image generation with clip latents,

    A. Ramesh, P. Dhariwal, A. Nichol, C. Chu, and M. Chen, “Hierarchical text-conditional image generation with clip latents,” 2022

    2022

  14. [22]

    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,” ICLR, 2024

    2024

  15. [23]

    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, ...

    2022

  16. [24]

    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

  17. [25]

    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, ...

    2022

  18. [26]

    Attend-and- excite: Attention-based semantic guidance for text-to-image diffusion models,

    H. Chefer, Y . Alaluf, Y . Vinker, L. Wolf, and D. Cohen-Or, “Attend-and- excite: Attention-based semantic guidance for text-to-image diffusion models,” 2023

    2023

  19. [27]

    Layerdiffusion: Layered controlled image editing with diffusion models,

    P. Li, Q. Huang, Y . Ding, and Z. Li, “Layerdiffusion: Layered controlled image editing with diffusion models,” SIGGRAPH Asia, 2023

    2023

  20. [28]

    Diffusion self-guidance for controllable image generation,

    D. Epstein, A. Jabri, B. Poole, A. A. Efros, and A. Holynski, “Diffusion self-guidance for controllable image generation,” NeurIPS, 2023

    2023

  21. [29]

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

    M. Cao, X. Wang, Z. Qi, Y . Shan, X. Qie, and Y . Zheng, “Masactrl: Tuning-free mutual self-attention control for consistent image synthesis and editing,” ICCV, 2023

    2023

  22. [30]

    Dragondiffusion: Enabling drag-style manipulation on diffusion models,

    C. Mou, X. Wang, J. Song, Y . Shan, and J. Zhang, “Dragondiffusion: Enabling drag-style manipulation on diffusion models,” ICLR, 2023

    2023

  23. [31]

    iedit: Localised text-guided image editing with weak supervision,

    R. Bodur, E. Gundogdu, B. Bhattarai, T.-K. Kim, M. Donoser, and L. Bazzani, “iedit: Localised text-guided image editing with weak supervision,” CVPRW, 2024

    2024

  24. [32]

    Sine: Single image editing with text-to-image diffusion models,

    Z. Zhang, L. Han, A. Ghosh, D. N. Metaxas, and J. Ren, “Sine: Single image editing with text-to-image diffusion models,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 6027–6037

    2023

  25. [33]

    Ledits: Real image editing with ddpm inversion and semantic guidance,

    L. Tsaban and A. Passos, “Ledits: Real image editing with ddpm inversion and semantic guidance,” arXiv preprint arXiv:2307.00522 , 2023

    2023 arXiv

  26. [34]

    Zone: Zero-shot instruction-guided local editing,

    S. Li, B. Zeng, Y . Feng, S. Gao, X. Liu, J. Liu, L. Li, X. Tang, Y . Hu, J. Liu et al., “Zone: Zero-shot instruction-guided local editing,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 6254–6263

    2024

  27. [35]

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

    L. Zhang and M. Agrawala, “Adding conditional control to text-to-image diffusion models,” ICCV, 2023

    2023

  28. [36]

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

    C. Mou, X. Wang, L. Xie, J. Zhang, Z. Qi, Y . Shan, and X. Qie, “T2i- adapter: Learning adapters to dig out more controllable ability for text- to-image diffusion models,” AAAI, 2024

    2024

  29. [37]

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

    S. Zhao, D. Chen, Y .-C. Chen, J. Bao, S. Hao, L. Yuan, and K.-Y . K. Wong, “Uni-controlnet: All-in-one control to text-to-image diffusion models,” NeurIPS, 2023

    2023

  30. [38]

    Directed diffusion: Direct control of object placement through attention guidance,

    W.-D. K. Ma, J. Lewis, W. B. Kleijn, and T. Leung, “Directed diffusion: Direct control of object placement through attention guidance,” arXiv preprint arXiv:2302.13153, 2023

    2023 arXiv

  31. [39]

    Ssr-encoder: Encoding selective subject represen- tation for subject-driven generation,

    Y . Zhang, Y . Song, J. Liu, R. Wang, J. Yu, H. Tang, H. Li, X. Tang, Y . Hu, H. Panet al., “Ssr-encoder: Encoding selective subject represen- tation for subject-driven generation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp...

    2024

  32. [40]

    LoRA: Low-rank adaptation of large language models,

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

    2022

  33. [41]

    Domain-agnostic tuning-encoder for fast personaliza- tion of text-to-image models,

    M. Arar, R. Gal, Y . Atzmon, G. Chechik, D. Cohen-Or, A. Shamir, and A. H. Bermano, “Domain-agnostic tuning-encoder for fast personaliza- tion of text-to-image models,” arXiv preprint arXiv:2307.06925 , 2023

    2023 arXiv

  34. [42]

    Taming encoder for zero fine-tuning image customization with text-to-image diffusion models,

    X. Jia, Y . Zhao, K. C. Chan, Y . Li, H. Zhang, B. Gong, T. Hou, H. Wang, and Y .-C. Su, “Taming encoder for zero fine-tuning image customization with text-to-image diffusion models,” arXiv preprint arXiv:2304.02642, 2023

    2023 arXiv

  35. [43]

    An image is worth one word: Personalizing text-to-image generation using textual inversion,

    R. Gal, Y . Alaluf, Y . Atzmon, O. Patashnik, A. H. Bermano, G. Chechik, and D. Cohen-Or, “An image is worth one word: Personalizing text-to-image generation using textual inversion,” 2022. [Online]. Available: https://arxiv.org/abs/2208.01618

    2022 arXiv

  36. [44]

    Hyperdreambooth: Hypernet- works for fast personalization of text-to-image models,

    N. Ruiz, Y . Li, V . Jampani, W. Wei, T. Hou, Y . Pritch, N. Wad- hwa, M. Rubinstein, and K. Aberman, “Hyperdreambooth: Hypernet- works for fast personalization of text-to-image models,” arXiv preprint arXiv:2307.06949, 2023

    2023 arXiv

  37. [45]

    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,” 2022

    2022

  38. [46]

    Fast personalized text to image synthesis with attention injection,

    Y . Zhang, Y . Song, J. Yu, H. Pan, and Z. Jing, “Fast personalized text to image synthesis with attention injection,” in ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2024, pp. 6195–6199

    2024

  39. [47]

    Ootdiffusion: Outfitting fusion based latent diffusion for controllable virtual try-on,

    Y . Xu, T. Gu, W. Chen, and C. Chen, “Ootdiffusion: Outfitting fusion based latent diffusion for controllable virtual try-on,” AAAI, 2025

    2025

  40. [48]

    Stablegarment: Garment-centric generation via stable diffusion,

    R. Wang, H. Guo, J. Liu, H. Li, H. Zhao, X. Tang, Y . Hu, H. Tang, and P. Li, “Stablegarment: Garment-centric generation via stable diffusion,” arXiv preprint arXiv:2403.10783 , 2024

    2024 arXiv

  41. [49]

    Stableviton: Learning semantic correspondence with latent diffusion model for virtual try-on,

    J. Kim, G. Gu, M. Park, S. Park, and J. Choo, “Stableviton: Learning semantic correspondence with latent diffusion model for virtual try-on,” CVPR, 2024

    2024

  42. [50]

    Cat-dm: Controllable accelerated virtual try-on with diffusion model,

    J. Zeng, D. Song, W. Nie, H. Tian, T. Wang, and A. Liu, “Cat-dm: Controllable accelerated virtual try-on with diffusion model,” CVPR, 2024

    2024

  43. [51]

    Stable-makeup: When real-world makeup transfer meets diffusion model,

    Y . Zhang, L. Wei, Q. Zhang, Y . Song, J. Liu, H. Li, X. Tang, Y . Hu, and H. Zhao, “Stable-makeup: When real-world makeup transfer meets diffusion model,” SIGGRAPH, 2025

    2025

  44. [52]

    Implicit neural representations with structured latent codes for human body modeling,

    S. Peng, C. Geng, Y . Zhang, Y . Xu, Q. Wang, Q. Shuai, X. Zhou, and H. Bao, “Implicit neural representations with structured latent codes for human body modeling,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023

    2023

  45. [53]

    Neural body: Implicit neural representations with structured latent codes for novel view synthesis of dynamic humans,

    S. Peng, Y . Zhang, Y . Xu, Q. Wang, Q. Shuai, H. Bao, and X. Zhou, “Neural body: Implicit neural representations with structured latent codes for novel view synthesis of dynamic humans,” in CVPR, 2021

    2021

  46. [54]

    Gaussianavatars: Photorealistic head avatars with rigged 3d gaussians,

    S. Qian, T. Kirschstein, L. Schoneveld, D. Davoli, S. Giebenhain, and M. Nießner, “Gaussianavatars: Photorealistic head avatars with rigged 3d gaussians,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 20 299–20 309

    2024

  47. [55]

    Nerf: Representing scenes as neural radiance fields for view synthesis,

    B. Mildenhall, P. P. Srinivasan, M. Tancik, J. T. Barron, R. Ramamoorthi, and R. Ng, “Nerf: Representing scenes as neural radiance fields for view synthesis,” in ECCV, 2020

    2020

  48. [56]

    3d gaussian splatting for real-time radiance field rendering,

    B. Kerbl, G. Kopanas, T. Leimk ¨uhler, and G. Drettakis, “3d gaussian splatting for real-time radiance field rendering,” ACM Transactions on Graphics , vol. 42, no. 4, July 2023. [Online]. Available: https://repo-sam.inria.fr/fungraph/3d-gaussian-splatting/

    2023

  49. [57]

    Zero-1-to-3: Zero-shot one image to 3d object,

    R. Liu, R. Wu, B. Van Hoorick, P. Tokmakov, S. Zakharov, and C. V on- drick, “Zero-1-to-3: Zero-shot one image to 3d object,” in Proceedings of the IEEE/CVF international conference on computer vision , 2023, pp. 9298–9309

    2023

  50. [58]

    Mvdream: Multi- view diffusion for 3d generation,

    Y . Shi, P. Wang, J. Ye, M. Long, K. Li, and X. Yang, “Mvdream: Multi- view diffusion for 3d generation,” ICLR, 2024

    2024

  51. [59]

    Zero123++: a single image to consistent multi-view diffusion base model,

    R. Shi, H. Chen, Z. Zhang, M. Liu, C. Xu, X. Wei, L. Chen, C. Zeng, and H. Su, “Zero123++: a single image to consistent multi-view diffusion base model,” arXiv preprint arXiv:2310.15110 , 2023

    2023 arXiv

  52. [60]

    Syncdreamer: Generating multiview-consistent images from a single- view image,

    Y . Liu, C. Lin, Z. Zeng, X. Long, L. Liu, T. Komura, and W. Wang, “Syncdreamer: Generating multiview-consistent images from a single- view image,” (ICLR), 2024

    2024

  53. [61]

    Objaverse: A universe of annotated 3d objects,

    M. Deitke, D. Schwenk, J. Salvador, L. Weihs, O. Michel, E. VanderBilt, L. Schmidt, K. Ehsani, A. Kembhavi, and A. Farhadi, “Objaverse: A universe of annotated 3d objects,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2023, pp. 13 142–13 153

    2023

  54. [62]

    Stable video diffusion: Scaling latent video diffusion models to large datasets,

    A. Blattmann, T. Dockhorn, S. Kulal, D. Mendelevitch, M. Kilian, D. Lorenz, Y . Levi, Z. English, V . V oleti, A. Lettset al., “Stable video diffusion: Scaling latent video diffusion models to large datasets,” arXiv preprint arXiv:2311.15127, 2023. 14

    2023 arXiv

  55. [63]

    Efficient geometry-aware 3D generative adversarial networks,

    E. R. Chan, C. Z. Lin, M. A. Chan, K. Nagano, B. Pan, S. D. Mello, O. Gallo, L. Guibas, J. Tremblay, S. Khamis, T. Karras, and G. Wet- zstein, “Efficient geometry-aware 3D generative adversarial networks,” in CVPR, 2022

    2022

  56. [64]

    Gram-hd: 3d-consistent image generation at high resolution with generative radiance manifolds,

    J. Xiang, J. Yang, Y . Deng, and X. Tong, “Gram-hd: 3d-consistent image generation at high resolution with generative radiance manifolds,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), October 2023, pp. 2195–2205

    2023

  57. [65]

    3d gan inversion with pose optimization,

    J. Ko, K. Cho, D. Choi, K. Ryoo, and S. Kim, “3d gan inversion with pose optimization,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision , 2023, pp. 2967–2976

    2023

  58. [66]

    Flashface: Human image personalization with high-fidelity identity preservation,

    S. Zhang, L. Huang, X. Chen, Y . Zhang, Z.-F. Wu, Y . Feng, W. Wang, Y . Shen, Y . Liu, and P. Luo, “Flashface: Human image personalization with high-fidelity identity preservation,” arXiv preprint arXiv:2403.17008, 2024

    2024

  59. [67]

    Animatediff: Animate your personalized text-to- image diffusion models without specific tuning,

    Y . Guo, C. Yang, A. Rao, Z. Liang, Y . Wang, Y . Qiao, M. Agrawala, D. Lin, and B. Dai, “Animatediff: Animate your personalized text-to- image diffusion models without specific tuning,” International Confer- ence on Learning Representations , 2024

    2024

  60. [68]

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

    M. Heusel, H. Ramsauer, T. Unterthiner, B. Nessler, and S. Hochreiter, “Gans trained by a two time-scale update rule converge to a local nash equilibrium,” Advances in neural information processing systems , vol. 30, 2017

    2017

  61. [69]

    Image quality assessment: from error visibility to structural similarity,

    Z. Wang, A. C. Bovik, H. R. Sheikh, and E. P. Simoncelli, “Image quality assessment: from error visibility to structural similarity,” IEEE transactions on image processing , vol. 13, no. 4, pp. 600–612, 2004

    2004

  62. [70]

    Arcface: Additive angular margin loss for deep face recognition,

    J. Deng, J. Guo, X. Niannan, and S. Zafeiriou, “Arcface: Additive angular margin loss for deep face recognition,” in CVPR, 2019

    2019

  63. [71]

    Learning transferable visual models from natural language supervi- sion,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark, G. Krueger, and I. Sutskever, “Learning transferable visual models from natural language supervi- sion,” 2021

    2021

  64. [72]

    Accurate 3d face reconstruction with weakly-supervised learning: From single image to image set,

    Y . Deng, J. Yang, S. Xu, D. Chen, Y . Jia, and X. Tong, “Accurate 3d face reconstruction with weakly-supervised learning: From single image to image set,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops , 2019, pp. 0–0

    2019

Pith tools

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