Pith. sign in

REVIEW 4 major objections 5 minor 53 references

Try-On-Adapter: A Simple and Flexible Try-On Paradigm

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Try-On-Adapter claims virtual try-on is an outpainting problem: given only a face image and a garment image, a diffusion model imagines the rest of the person and reports FID 5.56 paired and 7.23 unpaired on VITON-HD.

desk verdict Outpainting-based try-on is a real idea; the FID SOTA claim is not yet supported by the protocol-mismatched comparison. read the letter →

arxiv 2411.10187 v1 pith:S6M7NKH7 submitted 2024-11-15 cs.CV

Hanzhong Guo , Jianfeng Zhang , Cheng Zou , Jun Li , Meng Wang , Ruxue Wen , Pingzhong Tang , Jingdong Chen
show 1 more author
Ming Yang
This is my paper · ORCID
classification cs.CV
keywords virtualtry-onoutpaintingimage-as-promptlatentdiffusionTry-On-AdaptergarmentpreservationVITON-HDcontrollablegeneration
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

Try-On-Adapter (TOA) tries to establish that virtual try-on should be treated as outpainting, not inpainting. Where earlier systems require a clean full-body standing photo and only fill in the garment region, TOA takes a reference face and a garment image as its only visual inputs and lets the diffusion model imagine pose, body, and background. The paper argues this makes try-on simpler for users and more controllable, since text prompts and pose maps can steer the imagined body. On VITON-HD it reports the lowest FID among compared methods, 5.56 paired and 7.23 unpaired, while explicitly accepting lower SSIM and LPIPS than inpainting baselines because it regenerates more of the image. A sympathetic reading is that the point is practical generation from casual, low-quality inputs rather than pixel fidelity to a reference standing image.

What carries the argument

The mechanism that carries the argument is a pair of trainable adapter blocks wrapped around a frozen diffusion U-Net. The first, Comprehension- and Fusion-inspired Cross-attention, encodes the face and garment with CLIP's full hidden states, runs a small attention block over them with learned query, key, and value projections, and injects the result through decoupled cross-attention next to the text embedding. The second, Reference-preserving Self-attention, copies the denoising U-Net as a frozen Reference U-Net and concatenates its feature maps into the self-attention of each denoising layer, which is what keeps fine garment details from being lost. These adapters are trained with denoising score matching on a dataset reconstructed from public try-on images by SAM face crops and CogVLM captions, then fine-tuned in a second stage; ControlNet is added at inference for pose control.

What would settle it

Run TOA and the inpainting baselines under identical input conditions—either give all methods the same minimal face-plus-garment input, or force the same pose and background on every output—and recompute FID on VITON-HD; if TOA no longer beats StableVITON and OOTDiffusion, the central SOTA claim is an artifact of comparing different tasks.

Watch

Extended reading notes

Core claim

The central claim is that the try-on task can be reformulated as 'given a face and a garment, imagine the rest,' and that a diffusion model with two lightweight adapters is enough to do it convincingly. The Try-On-Adapter fuses CLIP encodings of the face and garment into the cross-attention layers of a frozen Stable Diffusion U-Net, alongside the text prompt, and uses a second Reference U-Net to feed garment appearance into self-attention so patterns and textures survive generation. Training data is built by cropping faces from standing images with SAM and captioning each image with CogVLM, then a two-stage schedule pretrains on web images and fine-tunes on VITON-HD. The paper reports that this recipe preserves the given face and garment, accepts text edits such as 'pink dress,' supports pose control through ControlNet, and can even consume a full standing image as its reference because random face-crop expansion makes that a special case.

Load-bearing premise

The load-bearing premise is that FID comparisons across the two paradigms are fair: TOA was free to choose pose, background, and body proportions to match the test distribution, whereas inpainting baselines had to stay close to a supplied standing image; if FID rewards that freedom, the reported advantage is protocol-driven rather than a sign of better rendering.

Editorial extensions

If this is right

  • A user would only need a selfie-style face crop and a product garment photo, not a clean full-body standing image, to get a realistic try-on.
  • Garment color and style become editable through text prompts, and human pose becomes controllable through ControlNet, in a way that inpainting try-on methods do not naturally support.
  • Because the model is free to generate pose, background, and body proportions, outputs can be more diverse than fidelity-to-input inpainting results.
  • Randomly enlarging the face crop lets the same trained model optionally accept a full standing reference image, making inpainting-style input a special case of the proposed paradigm.
  • Garment screenshots and lower-quality internet images remain usable, since the garment enters the model through CLIP-level features rather than pixel-aligned warping.

Reading between the lines

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

  • The paper leaves implicit that the fusion attention is not clothing-specific; the same face-plus-item recipe could plausibly extend to hats, bags, shoes, and other accessories, which the conclusion names as future work.
  • A fairer protocol test that the paper does not run would compare all methods on identical minimal input (face plus garment only), or with pose and background fixed; without that, the FID advantage is entangled with the outpainting model's freedom to invent the rest of the scene.
  • If the paradigm scales, the product-level consequence would be a shift in try-on from warping a given model photo toward synthesizing a new photo of the user from a phone selfie, with identity preservation as the main constraint.
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

4 major / 5 minor

Summary. The paper proposes Try-On-Adapter (TOA), a virtual try-on method that treats the task as outpainting rather than inpainting. Given only a cropped face image, a garment image, and optional text/pose conditioning, TOA generates a full dressed-person image using a Stable Diffusion backbone augmented with an IP-Adapter-style cross-attention block for face/garment embeddings and a Reference U-Net for garment detail preservation. The authors introduce a two-stage training pipeline (0.2M Internet images for pretraining, VITON-HD for finetuning) and report qualitative comparisons against OOTDiffusion, cross-dataset and low-quality-input experiments, and quantitative results on VITON-HD: FID 5.56 (paired) and 7.23 (unpaired), with SSIM 0.772 and LPIPS 0.178.

Significance. If the central claim were established, the paper would be practically significant: it would show that virtual try-on does not require a high-quality full standing image, and that pose and text editing can be combined with face/garment preservation in one diffusion-based system. The qualitative results support the feasibility of this outpainting paradigm, and the cross-dataset examples with internet garment screenshots are a useful demonstration of real-world applicability. However, the state-of-the-art quantitative claim is not yet load-bearing. The main FID comparison in Table 1 uses different input protocols for TOA and all baselines, and no matched-protocol or unconditional control is provided; the much lower SSIM/LPIPS are attributed to the paradigm but not demonstrated as inherent to outpainting. I also note that the manuscript includes no implementation details, no error bars, and no public code/checkpoints, so reproducibility cannot currently be assessed.

major comments (4)
  1. [Quantitative Results / Table 1] The headline FID comparison is not a fair protocol comparison. Every inpainting baseline receives a full standing person image plus a garment, whereas TOA receives only a face crop and a garment. Because FID measures the distributional match between generated and real VITON-HD images, TOA is free to choose pose, body shape, and background composition to match the dataset prior, while the inpainting baselines are constrained to preserve all unmasked pixels of a fixed input. The reported FID advantage (5.56 paired, 7.23 unpaired) may therefore reflect a difference in input freedom rather than better try-on quality. The authors attribute this to the paradigm, but they do not provide the necessary control: an unconditional or face/garment-only baseline, or a matched-input version of an inpainting method. Without such a control, the state-of-the-art claim is not established.
  2. [Quantitative Results / Paired Evaluation] The paper's own SSIM (0.772 vs 0.862–0.877) and LPIPS (0.178 vs 0.071–0.117) show much weaker pixel-level and perceptual similarity to the ground truth, and the text states this is 'mainly due to the difference of paradigms'. This attribution needs quantitative support rather than assertion. The manuscript should either (a) feed the same face-only + garment protocol to a baseline, (b) feed a full standing image to TOA (the paper says the randomly-expanded face can cover the whole image), or (c) provide an ablation that isolates the amount of preserved context. As it stands, the low SSIM/LPIPS could indicate weaker conditioning adherence rather than an inherent property of outpainting.
  3. [Training and Inference / Experimental settings] The paper's quantitative results are not reproducible or fully auditable as submitted. The training schedule relies on 0.2M Internet images, two stages, a random face-expansion range, a classifier-free guidance drop rate, and face/garment extraction with SAM, but the manuscript defers all concrete hyperparameters, dataset sizes, and evaluation details to an Appendix that is not included. Furthermore, FID scores are reported without error bars or multiple seeds; on a 6K test set, the difference between TOA and the next-best FID may be within sampling noise. Please report variance estimates and a detailed implementation description.
  4. [Qualitative comparisons] The qualitative comparison with OOTDiffusion in Fig. 4 is made under different input modalities (full standing image for OOTDiffusion vs face+garment for TOA) and is evaluated only subjectively. This is a reasonable illustrative comparison, but it should not be extended into a claim of better performance. A user study with side-by-side images or a matched-input quantitative evaluation is needed before the qualitative advantage is used as evidence for the SOTA claim.
minor comments (5)
  1. [Single Dataset Evaluation / Fig. 3 caption] The caption contains the typo 'VITION-HD'; it should read 'VITON-HD'.
  2. [Eq. (4)] The notation 'Ez0,y∼[ε(q0),y]Eq(ϵ)' is malformed; please clarify the joint distribution over z0 and y.
  3. [Table 1] The first method row is labelled 'VITON-HD (Choi et al. 2021)', which is the dataset name rather than a method name; rename it to the actual baseline method to avoid confusion.
  4. [Constructing the Training Data] The claim that traditional methods are 'just a special case' of TOA because the face region can be expanded to the whole image is not fully supported: a whole standing image reference contains body, background, and garment-pose information that the model is expected to preserve rather than imagine, so the conditional task changes in kind, not just in input size. Please clarify what exactly is recovered when the face box is expanded to the full image.
  5. [Diffusion Models background] Equation (2) has a likely formatting error in the noise-scale term ('1 − α2(t)p' should presumably be '1 − α2(t)'); please correct the typesetting.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the outpainting formulation is an empirical design choice, and the FID claims are external-benchmark measurements, not consequences of fitted inputs or self-citations.

full rationale

The paper's core claim is that virtual try-on can be reformulated as outpainting and that the resulting TOA model achieves strong FID on VITON-HD. There is no derivation chain in which a predicted quantity equals an input by construction. Eq. (7) is a standard denoising score-matching objective conditioned on face and garment embeddings produced by adapters; the reported FID values are computed on the held-out VITON-HD test set after two-stage training, and no parameter is fitted to the FID itself or to a subset of the test set. The lower SSIM and higher LPIPS are explicitly attributed to the paradigm difference, not hidden as successes, and the FID advantage is an empirical measurement rather than a reduced-form identity. The cited components (IP-Adapter, InstantID, AnimateAnyone, CLIP, Stable Diffusion) are external prior work and are not invoked as a self-citation chain; no uniqueness theorem from the present authors is used to force the design. The evaluation protocol difference—TOA receives only a face and garment while inpainting baselines receive a full standing image—is a legitimate validity concern about metric fairness, but it is not circular reasoning under the definitions used here, because the FID comparison is not logically entailed by the training objective or by any fitted parameter. Overall, the central claim has independent empirical content and the manuscript is self-contained against an external benchmark, so the circularity score is 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

No new physical or theoretical entities are introduced; the components are adapters on top of existing diffusion models.

free parameters (3)
  • Random face expansion range = not specified
    The paper says it randomly expands the SAM face area, with larger expansions approaching the whole image. The chosen range is a hand-tuned parameter that controls how much body context leaks into the face condition and is central to the outpainting comparison.
  • Classifier-free guidance drop rate = not specified
    The paper randomly drops the image embedding and replaces it with zero embeddings during training; the drop probability is not reported and directly affects generation quality and diversity.
  • Two-stage data scale = 0.2M internet images plus VITON-HD fine-tune
    The pretraining corpus size and fine-tuning schedule are design choices selected by the authors; different scales likely change the reported FID.
assumptions (4)
  • domain assumption A pre-trained Stable Diffusion U-Net retains rich priors for color, texture, and pose that are sufficient for try-on after adapter training.
    The whole method relies on transferring Stable Diffusion priors to the try-on task without retraining the base network.
  • domain assumption CLIP's image embeddings preserve the garment details needed for try-on.
    The design chooses CLIP as the image encoder because it handles color and texture; if CLIP loses fine patterns at 224x224, garment fidelity depends on the Reference U-Net to compensate.
  • domain assumption SAM face crops and CogVLM captions are accurate enough to serve as training conditions.
    The training set is built by automatically segmenting faces and generating captions; errors in these tools propagate into the learned adapter.
  • ad hoc to paper VITON-HD is a valid benchmark for the new outpainting protocol.
    The paper evaluates an outpainting method on a dataset designed for inpainting, and no outpainting-specific benchmark or ground-truth protocol is introduced.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Try-On-Adapter: A Simple and Flexible Try-On Paradigm." pith.science (2026). https://pith.science/paper/S6M7NKH7

@misc{pith2026241110187,
  author       = {Pith},
  title        = {Pith review of: Try-On-Adapter: A Simple and Flexible Try-On Paradigm},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S6M7NKH7}},
  note         = {Machine review of arXiv:2411.10187}
}
read the original abstract

Image-based virtual try-on, widely used in online shopping, aims to generate images of a naturally dressed person conditioned on certain garments, providing significant research and commercial potential. A key challenge of try-on is to generate realistic images of the model wearing the garments while preserving the details of the garments. Previous methods focus on masking certain parts of the original model's standing image, and then inpainting on masked areas to generate realistic images of the model wearing corresponding reference garments, which treat the try-on task as an inpainting task. However, such implements require the user to provide a complete, high-quality standing image, which is user-unfriendly in practical applications. In this paper, we propose Try-On-Adapter (TOA), an outpainting paradigm that differs from the existing inpainting paradigm. Our TOA can preserve the given face and garment, naturally imagine the rest parts of the image, and provide flexible control ability with various conditions, e.g., garment properties and human pose. In the experiments, TOA shows excellent performance on the virtual try-on task even given relatively low-quality face and garment images in qualitative comparisons. Additionally, TOA achieves the state-of-the-art performance of FID scores 5.56 and 7.23 for paired and unpaired on the VITON-HD dataset in quantitative comparisons.

Figures

Figures reproduced from arXiv: 2411.10187 by the authors.

Figure 1
Figure 1. Results of Try-On-Adapter. The first column gives reference face (top) and reference pose (bottom). The second column shows two target garments. The third to fifth columns show try-on results generated by TOA with different conditions. The third column is generated by the reference face and target garments, conditioned on null text guidance, without pose control. The fourth column is generated by the reference face … view at source ↗
Figure 2
Figure 2. Overall architecture of Try-On-Adapter. The Blocks in red represent the trainable blocks while the blocks in blue denote the frozen pretrained blocks. The blocks in orange represent the original available public try-on dataset and the blocks in yellow represent the obtained dataset. The proposed Try-On-Adapter primarily comprises two modules: the fusion of different references on the Cross-Attention Block and the Re… view at source ↗
Figure 3
Figure 3. Qualitative Results of the Try-On-Adapter. (a) is the single dataset evaluation where the faces and the garments are from the VITON-HD and (b) denotes the cross dataset evaluation where the faces are from the VITON-HD and the garments are from the Internet. (a)-(2) denotes the generation of TOA while given the pose in (a)-(1) and the face in (a)-(4). (I)-(a)-(3) denotes the generation under text guidance ”pink dress… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 14 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Bao, F.; Nie, S.; Xue, K.; Li, C.; Pu, S.; Wang, Y.; Yue, G.; Cao, Y.; Su, H.; and Zhu, J. 2023. One Transformer Fits All Distributions in Multi-Modal Diffusion at Scale. arXiv preprint arXiv:2303.06555

    arXiv 2023

  4. [4]

    Betker, J.; Goh, G.; Jing, L.; Brooks, T.; Wang, J.; Li, L.; Ouyang, L.; Zhuang, J.; Lee, J.; Guo, Y.; et al. 2023. Improving image generation with better captions. Computer Science. https://cdn. openai. com/papers/dall-e-3. pdf, 2(3): 8

    2023

  5. [5]

    Choi, S.; Park, S.; Lee, M.; and Choo, J. 2021. Viton-hd: High-resolution virtual try-on via misalignment-aware normalization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 14131--14140

    2021

  6. [6]

    Civitai. 2020. https://https://civitai.com/

    work page 2020

  7. [7]

    Cui, A.; Mahajan, J.; Shah, V.; Gomathinayagam, P.; and Lazebnik, S. 2023. Street TryOn: Learning In-the-Wild Virtual Try-On from Unpaired Person Images. arXiv preprint arXiv:2311.16094

    arXiv 2023

  8. [8]

    Dhariwal, P.; and Nichol, A. 2021. Diffusion models beat gans on image synthesis. Advances in Neural Information Processing Systems, 34: 8780--8794

    2021

Show all 53 references
  1. [9]

    Duchon, J. 1977. Splines minimizing rotation-invariant semi-norms in Sobolev spaces. In Constructive Theory of Functions of Several Variables: Proceedings of a Conference Held at Oberwolfach April 25--May 1, 1976, 85--100. Springer

    1977

  2. [10]

    Ge, C.; Song, Y.; Ge, Y.; Yang, H.; Liu, W.; and Luo, P. 2021 a . Disentangled cycle consistency for highly-realistic virtual try-on. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 16928--16937

    2021

  3. [11]

    Ge, Y.; Song, Y.; Zhang, R.; Ge, C.; Liu, W.; and Luo, P. 2021 b . Parser-free virtual try-on via distilling appearance flows. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 8485--8493

    2021

  4. [12]

    Gong, K.; Liang, X.; Li, Y.; Chen, Y.; Yang, M.; and Lin, L. 2018. Instance-level human parsing via part grouping network. In Proceedings of the European conference on computer vision (ECCV), 770--785

    2018

  5. [13]

    A.; Neverova, N.; and Kokkinos, I

    G \"u ler, R. A.; Neverova, N.; and Kokkinos, I. 2018. Densepose: Dense human pose estimation in the wild. In Proceedings of the IEEE conference on computer vision and pattern recognition, 7297--7306

    2018

  6. [14]

    Han, X.; Hu, X.; Huang, W.; and Scott, M. R. 2019. Clothflow: A flow-based model for clothed person generation. In Proceedings of the IEEE/CVF international conference on computer vision, 10471--10480

    2019

  7. [15]

    Han, X.; Wu, Z.; Wu, Z.; Yu, R.; and Davis, L. S. 2018. Viton: An image-based virtual try-on network. In Proceedings of the IEEE conference on computer vision and pattern recognition, 7543--7552

    2018

  8. [16]

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

    2017

  9. [17]

    Ho, J.; Jain, A.; and Abbeel, P. 2020. Denoising diffusion probabilistic models. Advances in Neural Information Processing Systems, 33: 6840--6851

    2020

  10. [18]

    J.; Norouzi, M.; and Salimans, T

    Ho, J.; Saharia, C.; Chan, W.; Fleet, D. J.; Norouzi, M.; and Salimans, T. 2022 a . Cascaded Diffusion Models for High Fidelity Image Generation. J. Mach. Learn. Res., 23(47): 1--33

    2022

  11. [19]

    Ho, J.; Salimans, T.; Gritsenko, A.; Chan, W.; Norouzi, M.; and Fleet, D. J. 2022 b . Video diffusion models. arXiv preprint arXiv:2204.03458

    2022 arXiv

  12. [20]

    C.; Su, Y.-C.; Chen, W.; Li, Y.; Sohn, K.; Zhao, Y.; Ben, X.; Gong, B.; Cohen, W.; et al

    Hu, H.; Chan, K. C.; Su, Y.-C.; Chen, W.; Li, Y.; Sohn, K.; Zhao, Y.; Ben, X.; Gong, B.; Cohen, W.; et al. 2024. Instruct-Imagen: Image generation with multi-modal instruction. arXiv preprint arXiv:2401.01952

    2024 arXiv

  13. [21]

    Hu, L.; Gao, X.; Zhang, P.; Sun, K.; Zhang, B.; and Bo, L. 2023. Animate anyone: Consistent and controllable image-to-video synthesis for character animation. arXiv preprint arXiv:2311.17117

    2023 arXiv

  14. [22]

    Jaderberg, M.; Simonyan, K.; Zisserman, A.; et al. 2015. Spatial transformer networks. Advances in neural information processing systems, 28

    2015

  15. [23]

    Karras, T.; Aittala, M.; Aila, T.; and Laine, S. 2022. Elucidating the design space of diffusion-based generative models. arXiv preprint arXiv:2206.00364

    2022 arXiv

  16. [24]

    Kim, J.; Gu, G.; Park, M.; Park, S.; and Choo, J. 2023. StableVITON: Learning Semantic Correspondence with Latent Diffusion Model for Virtual Try-On. arXiv preprint arXiv:2312.01725

    2023 arXiv

  17. [25]

    C.; Lo, W.-Y.; et al

    Kirillov, A.; Mintun, E.; Ravi, N.; Mao, H.; Rolland, C.; Gustafson, L.; Xiao, T.; Whitehead, S.; Berg, A. C.; Lo, W.-Y.; et al. 2023. Segment anything. arXiv preprint arXiv:2304.02643

    2023 arXiv

  18. [26]

    Kong, Z.; Ping, W.; Huang, J.; Zhao, K.; and Catanzaro, B. 2020. Diffwave: A versatile diffusion model for audio synthesis. arXiv preprint arXiv:2009.09761

    2020 arXiv

  19. [27]

    M.; Varadharajan, S.; and Kemelmacher-Shlizerman, I

    Lewis, K. M.; Varadharajan, S.; and Kemelmacher-Shlizerman, I. 2021. Tryongan: Body-aware try-on via layered interpolation. ACM Transactions on Graphics (TOG), 40(4): 1--10

    2021

  20. [28]

    Li, D.; Li, J.; and Hoi, S. 2024. Blip-diffusion: Pre-trained subject representation for controllable text-to-image generation and editing. Advances in Neural Information Processing Systems, 36

    2024

  21. [29]

    Li, Y.; Huang, C.; and Loy, C. C. 2019. Dense intrinsic appearance flow for human pose transfer. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 3693--3702

    2019

  22. [30]

    Li, Z.; Cao, M.; Wang, X.; Qi, Z.; Cheng, M.-M.; and Shan, Y. 2023. Photomaker: Customizing realistic human photos via stacked id embedding. arXiv preprint arXiv:2312.04461

    2023 arXiv

  23. [31]

    Ma, L.; Jia, X.; Sun, Q.; Schiele, B.; Tuytelaars, T.; and Van Gool, L. 2017. Pose guided person image generation. Advances in neural information processing systems, 30

    2017

  24. [32]

    Pan, X.; Dong, L.; Huang, S.; Peng, Z.; Chen, W.; and Wei, F. 2024. Generating Images in Context with Multimodal Large Language Models. In The Twelfth International Conference on Learning Representations

    2024

  25. [33]

    Podell, D.; English, Z.; Lacey, K.; Blattmann, A.; Dockhorn, T.; M \"u ller, J.; Penna, J.; and Rombach, R. 2023. Sdxl: Improving latent diffusion models for high-resolution image synthesis. arXiv preprint arXiv:2307.01952

    2023 arXiv

  26. [34]

    T.; and Mildenhall, B

    Poole, B.; Jain, A.; Barron, J. T.; and Mildenhall, B. 2022. Dreamfusion: Text-to-3d using 2d diffusion. arXiv preprint arXiv:2209.14988

    2022 arXiv

  27. [35]

    Popov, V.; Vovk, I.; Gogoryan, V.; Sadekova, T.; and Kudinov, M. 2021. Grad-tts: A diffusion probabilistic model for text-to-speech. In International Conference on Machine Learning, 8599--8608. PMLR

    2021

  28. [36]

    W.; Hallacy, C.; Ramesh, A.; Goh, G.; Agarwal, S.; Sastry, G.; Askell, A.; Mishkin, P.; Clark, J.; et al

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

    2021

  29. [37]

    Rombach, R.; Blattmann, A.; Lorenz, D.; Esser, P.; and Ommer, B. 2022. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 10684--10695

    2022

  30. [38]

    Sohl-Dickstein, J.; Weiss, E.; Maheswaranathan, N.; and Ganguli, S. 2015. Deep unsupervised learning using nonequilibrium thermodynamics. In International Conference on Machine Learning, 2256--2265. PMLR

    2015

  31. [39]

    P.; Kumar, A.; Ermon, S.; and Poole, B

    Song, Y.; Sohl-Dickstein, J.; Kingma, D. P.; Kumar, A.; Ermon, S.; and Poole, B. 2020. Score-based generative modeling through stochastic differential equations. arXiv preprint arXiv:2011.13456

    2020 arXiv

  32. [40]

    N.; Kaiser, .; and Polosukhin, I

    Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A. N.; Kaiser, .; and Polosukhin, I. 2017. Attention is all you need. Advances in neural information processing systems, 30

    2017

  33. [41]

    Wang, B.; Zheng, H.; Liang, X.; Chen, Y.; Lin, L.; and Yang, M. 2018. Toward characteristic-preserving image-based virtual try-on network. In Proceedings of the European conference on computer vision (ECCV), 589--604

    2018

  34. [42]

    Wang, Q.; Bai, X.; Wang, H.; Qin, Z.; and Chen, A. 2024 a . InstantID: Zero-shot Identity-Preserving Generation in Seconds. arXiv preprint arXiv:2401.07519

    2024 arXiv

  35. [43]

    Z.; Khabsa, M.; Fang, H.; and Ma, H

    Wang, S.; Li, B. Z.; Khabsa, M.; Fang, H.; and Ma, H. 2020. Linformer: Self-attention with linear complexity. arXiv preprint arXiv:2006.04768

    2020 arXiv

  36. [44]

    Wang, W.; Lv, Q.; Yu, W.; Hong, W.; Qi, J.; Wang, Y.; Ji, J.; Yang, Z.; Zhao, L.; Song, X.; et al. 2023. Cogvlm: Visual expert for pretrained language models. arXiv preprint arXiv:2311.03079

    2023 arXiv

  37. [45]

    C.; Sheikh, H

    Wang, Z.; Bovik, A. C.; Sheikh, H. R.; and Simoncelli, E. P. 2004. Image quality assessment: from error visibility to structural similarity. IEEE transactions on image processing, 13(4): 600--612

    2004

  38. [46]

    Wang, Z.; Lu, C.; Wang, Y.; Bao, F.; Li, C.; Su, H.; and Zhu, J. 2024 b . Prolificdreamer: High-fidelity and diverse text-to-3d generation with variational score distillation. Advances in Neural Information Processing Systems, 36

    2024

  39. [47]

    Xu, X.; Wang, Z.; Zhang, E.; Wang, K.; and Shi, H. 2022. Versatile Diffusion: Text, Images and Variations All in One Diffusion Model. arXiv preprint arXiv:2211.08332

    2022 arXiv

  40. [48]

    Xu, Y.; Gu, T.; Chen, W.; and Chen, C. 2024. OOTDiffusion: Outfitting Fusion based Latent Diffusion for Controllable Virtual Try-on. arXiv:2403.01779

    2024 arXiv

  41. [49]

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

    2023 arXiv

  42. [50]

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

    2023

  43. [51]

    A.; Shechtman, E.; and Wang, O

    Zhang, R.; Isola, P.; Efros, A. A.; Shechtman, E.; and Wang, O. 2018. The unreasonable effectiveness of deep features as a perceptual metric. In Proceedings of the IEEE conference on computer vision and pattern recognition, 586--595

    2018

  44. [52]

    Zhang, R.; Yang, W.; Peng, Z.; Wei, P.; Wang, X.; and Lin, L. 2019. Progressively diffused networks for semantic visual parsing. Pattern Recognition, 90: 78--86

    2019

  45. [53]

    Zhu, L.; Yang, D.; Zhu, T.; Reda, F.; Chan, W.; Saharia, C.; Norouzi, M.; and Kemelmacher-Shlizerman, I. 2023. TryOnDiffusion: A Tale of Two UNets. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 4606--4615

    2023

Pith tools

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