Learning Implicit Features with Flow Infused Attention for Realistic Virtual Try-On

Delong Zhang; Pengfei Xiong; Qiwei Huang; Wei-Shi Zheng; Wei Zhang; Yang Sun; Yuanliu Liu

REVIEW 3 major objections 6 minor 1 cited by

Learning Implicit Features with Flow Infused Attention for Realistic Virtual Try-On

T0 review · 3 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The central claim is that a dense garment-to-model warp flow, injected through cross-attention as a soft offset to diffusion features, gives virtual try-on both accurate pose fitting and fine texture preservation, beating explicit-warping…

desk verdict Clever flow-infused attention for try-on, with a solid DressCode showing, but the VITON-HD SOTA claim is undercut by a resampling inconsistency. read the letter →

arxiv 2412.11435 v1 pith:63RULTRN submitted 2024-12-16 cs.CV

Delong Zhang , Qiwei Huang , Yuanliu Liu , Yang Sun , Wei-Shi Zheng , Pengfei Xiong , Wei Zhang This is my paper · ORCID

classification cs.CV

keywords virtualtry-ondiffusionmodelimplicitwarpingflow-infusedattentioncross-attentiongarmentdetailpreservationStableFashionCLIP

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

Image-based virtual try-on faces a trade-off between fitting a garment to a new pose and preserving its original texture. FIA-VTON's proposal is to stop choosing: it computes a dense warp flow between the garment and the target person and injects that flow into a Stable Diffusion inpainting UNet through a Flow Infused Attention module, so the flow guides deformation softly instead of being applied as a hard warp. The paper reports state-of-the-art scores on VITON-HD and DressCode, improving alignment metrics (SSIM, PSNR, LPIPS) and distribution metrics (FID) over warp-based and learning-based baselines. If the claim holds, virtual try-on quality no longer needs to hinge on the precision of an explicit warping module.

What carries the argument

The central object is the Flow Infused Attention (FIA) module, which replaces every cross-attention layer of the denoising UNet. It takes a dense warp flow $F \in \mathbb{R}^{h \times w \times 2}$, a local garment feature $G$, the upstream model feature $P$, and a FashionCLIP spatial embedding $S$. The flow is projected by MLPs and added to $P$ and $G$, the summed features go through joint cross-attention with flow-modulated keys and values, and the result is modulated a second time by cross-attention with $S$. This lets the flow act as a learned offset on features rather than a hard warp on pixels, so the diffusion model can correct flow errors while still following the deformation pattern.

What would settle it

Rerun every method on VITON-HD under one common downsampling protocol and recompute FIDp and FIDu; if FIA-VTON's unpaired FID is not below GarDiff's 7.89, the headline state-of-the-art claim rests on the resampling difference the paper itself flags.

Watch

Extended reading notes

Core claim

FIA-VTON claims that the dense warp flow from garment to target model, when projected into the feature space of a Stable Diffusion inpainting UNet and injected through cross-attention together with local garment features and high-level FashionCLIP spatial features, guides the model to deform garments to the target pose while preserving fine texture. The paper reports that this implicit flow guidance outperforms explicit-warping pipelines such as VITON-HD, HR-VTON, GP-VTON, DCI-VTON, and D4-VTON, and learning-based diffusion methods such as LaDI-VTON, StableVITON, IDM-VTON, and GarDiff, with the largest gains on paired FID and SSIM. Ablations show that removing either the Flow Guider or the Spatial Guider degrades every metric, and that replacing the flow-attention fusion with concatenation or pixel-wise multiplication of the warped garment harms detail and fit.

Load-bearing premise

The state-of-the-art claim stands on the assumption that FID scores computed with different downsampling choices are comparable, even though the paper reports that bilinear versus cubic downsampling changes FID scores enough to move which variant appears best.

Editorial extensions

If this is right

Virtual try-on quality becomes less dependent on the exact accuracy of the warping module, because flow errors become soft corrections instead of hard pixel distortions.
The warped-garment plus inpainting pipeline can be replaced by a single diffusion UNet that receives flow, local garment features, and spatial features through one attention mechanism.
DressCode results across dresses, lower garments, and upper garments indicate the mechanism transfers across garment categories and handles semi-transparent materials.
Because the Flow Guider is frozen during training, FIA learns to exploit an off-the-shelf dense flow estimator without additional flow supervision.
Flow-guided implicit attention could be adopted by other diffusion-based editing tasks where a correspondence field between two images is known.

Reading between the lines

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

The paper's headline unpaired FID on VITON-HD (7.869) is computed with cubic downsampling, while the baselines and the default model use bilinear; because the paper itself says downsampling changes FID substantially, a reader should treat the bilinear default (7.914) as the comparable number, and that default is close to GarDiff's 7.89.
Making the flow estimator trainable end-to-end with FIA is a natural next step; since flow quality is load-bearing for detail reconstruction, joint training could produce larger gains than fixing the DSDM flow module.
The flow-as-soft-offset idea should transfer to other image editing tasks with known correspondence fields, such as pose transfer, expression reenactment, or relighting, where explicit warping is also brittle.

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.

Desk Editor's Note

read the letter

Colleague, quick take on arXiv:2412.11435 (FIA-VTON). The core idea is genuinely new: instead of warping the garment explicitly or feeding the warped garment to the diffusion model, they take a dense flow map from D4-VTON's DSDM, project it into the query/key space of cross-attention as additive offsets, and combine that with a FashionCLIP spatial stream. The decoupled cross-attention design is clean, and the ablations in Table 3 show both the flow and spatial modules contribute, with the biggest FIDp drop from removing the flow guider. On DressCode, the method beats GarDiff and D4-VTON consistently across all metrics. That's a solid result.

The soft spot is the VITON-HD headline. The paper itself states (Sec 4.1) that downsampling method significantly affects FID and that bilinear is the default, following GP-VTON and D4-VTON. Yet Table 1 bolds Ours* FIDu=7.869 from a cubic variant, while the default Ours is 7.914, slightly worse than GarDiff's 7.89. So the abstract's 'significantly outperform' claim doesn't hold for the default protocol. Section 4.2 also says 'best score of 7.869' under the FIDp metric, but 7.869 is the FIDu value. That's a mislabeling, likely a typo, but it's sloppy. The paired FIDp and SSIM/PSNR do favor Ours on VITON-HD, so it's not all bad, but the SOTA comparison needs fixing.

Also, no code or checkpoints are released, and baseline numbers are quoted from papers rather than rerun. That's not unusual for this area, but it means the central claim isn't independently verifiable right now. The frozen DSDM flow guider is a reasonable choice, and the FashionCLIP spatial stream is a nice complement to the Garment Net.

If I were refereeing, I'd ask for a corrected table with comparable resampling, a clarification of the metric labeling, and a code release. The architecture is worth engaging with; the DressCode results alone would justify that. It deserves a serious referee, not a desk reject. I'd take it to reading group if you want to see how a plausible conditioning mechanism can be undermined by evaluation protocol details.

Referee Report

3 major / 6 minor

Summary. FIA-VTON proposes a diffusion-based virtual try-on method in which a precomputed dense garment-to-model warp flow, local garment features from a garment UNet, and high-level spatial features from FashionCLIP are jointly injected into the cross-attention layers of a Stable Diffusion inpainting UNet. The flow is added to the query and key features rather than being used to explicitly warp the garment, and the authors argue this implicit guidance is less sensitive to flow estimation errors while still preserving detail. The method is evaluated on VITON-HD and DressCode with paired and unpaired FID, SSIM, PSNR, and LPIPS, and the paper claims state-of-the-art results in both datasets.

Significance. The core idea is timely and plausible: conditioning a diffusion try-on model with a dense warp flow through attention, rather than through an explicit warped garment, is a meaningful design choice and the ablation study supports the contribution of both the Flow Guider and the Spatial Guider. If the quantitative claims were robust, the paper would provide a useful alternative to explicit-warping pipelines. The authors also deserve credit for explicitly acknowledging in Section 4.1 that the image downsampling method has a significant impact on FID scores. However, the headline SOTA claim is not currently supported in a load-bearing way: on VITON-HD, the default bilinear model has FIDu 7.914, which is worse than GarDiff's 7.89, and LPIPS 0.047 is worse than GarDiff's 0.036. The bolded FIDu value 7.869 comes from a cubic-downsampling variant, a protocol not shared by any baseline. The DressCode results are more consistently in the paper's favor, but the same protocol concern applies there because the baselines are quoted rather than recomputed. The contribution is worth publishing only after the evaluation protocol is fixed or the claims are recalibrated.

major comments (3)

[§4.1, Table 1] The headline FIDu comparison is not valid under the paper's own protocol. The Table 1 caption states that all numbers use bilinear downsampling except Ours*, which uses cubic downsampling, and Section 4.1 states that different downsampling methods have a significant impact on FID scores. Therefore the bold FIDu value of 7.869 for Ours* cannot be fairly compared with the bilinear baseline FIDu values. Under the default bilinear setting, Ours achieves FIDu 7.914, which is worse than GarDiff's 7.89. The abstract's claim of significantly outperforming state-of-the-art methods is not supported for FIDu on VITON-HD; please recompute all baselines under the same resampling protocol, or remove the SOTA claim for this metric.
[§4.2, VITON-HD paragraph] The text says 'our method achieves outstanding performance in the FID p metric, with a best score of 7.869,' but 7.869 is the FIDu value of Ours*, not the FIDp value; the FIDp value of the default Ours is 4.686. Moreover, on VITON-HD the default model's LPIPS is 0.047, worse than GarDiff's 0.036, and its FIDu is not the best; only FIDp and a marginal SSIM difference (0.913 vs. 0.912) favor the method. The claim of consistent, significant superiority on VITON-HD should be softened or substantiated with a direct comparison using the default configuration.
[Table 2 and §4.1] For DressCode, the paper does not state which downsampling protocol was used for the reported FID numbers, nor whether the baseline FID scores were recomputed by the authors or quoted from other papers. Given the documented sensitivity of FID to downsampling in Section 4.1, the DressCode comparisons need the same protocol specification before the DressCode SOTA claim can be accepted. Please state the protocol and, ideally, rerun the baselines or report them with the exact same preprocessing.

minor comments (6)

[Throughout] There are several typos and wording issues, including 'Wrap Guidance' in the Related Work heading, 'LdDI-VTON' in Section 4.2, 'Spatil Guider' in Section 4.3, 'senarios' in the Section 4.4 title, and 'Table Table 1' in Section 4.2.
[Table 3] The table title says 'Ablation study on each module of FIA-Diff', but the method is called FIA-VTON throughout the rest of the paper; please make the names consistent.
[Eq. (5)] Equation (5) contains a trailing comma in the argument list of epsilon_theta: `∥ϵθ(xt, xg, F, S,) − ϵ∥`; this should be cleaned up.
[§3.3] FashionCLIP is used as the Spatial Guider, but no reference or model version is given; please cite the source and specify the exact feature-map resolution used as the 'highest level feature map without visual projection'.
[§4.3] The ablation text lists replacing FashionCLIP with OpenCLIP as one setting, but the body never discusses the corresponding Table 3 row ('w/ openclip'); either add the discussion or remove the row.
[§3.1] The phrase 'There several possible ways' is grammatically incomplete, and the discussion of the sparsity of warp flow could be more precise about how summation is more stable than multiplication for sparse two-dimensional offsets.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: FIA-VTON's flow-guided attention is an empirical architecture trained with a standard denoising loss on held-out test splits, with no fitted parameter or self-citation chain doing the work of the claimed result.

full rationale

The paper's derivation chain is not circular. The central claim is that feeding a dense warp flow (from the frozen DSDM module of D4-VTON), a local garment feature, and a FashionCLIP spatial feature into decoupled cross-attention layers of a Stable Diffusion inpainting UNet yields state-of-the-art virtual try-on. This is an empirical architecture claim, not a mathematical derivation from fitted constants. The only learned parameters are optimized with the standard denoising loss (Eq. 5) on training pairs, and the evaluation metrics (SSIM, PSNR, LPIPS, FID) are computed on held-out test splits of VITON-HD and DressCode. The dense flow map is produced by a fixed external network and is not a fitted parameter renamed as a prediction; the Spatial Guider is a pretrained external FashionCLIP model, not a hidden version of the target output. There are no self-citations by the present authors in the reference list, and the cited components (D4-VTON, GP-VTON, GarDiff, TryOnDiffusion) are external prior work used as baselines or modules, so no load-bearing self-citation chain exists. The most notable weaknesses are evaluation-protocol issues, not circularity: the paper states in Sec. 4.1 that 'different downsampling methods have a significant impact on both paired and unpaired FID scores,' and Table 1 shows the bold FIDu value 7.869 is from 'Ours*' with cubic downsampling while baselines use bilinear, with the default 'Ours' FIDu being 7.914. Additionally, Sec. 4.2 says 'our method achieves outstanding performance in the FID p metric, with a best score of 7.869,' although 7.869 is the FIDu value for Ours*. These are correctness and comparability concerns about the SOTA claim, but they do not make the derivation equivalent to its inputs by construction. No circular step is exhibited, so the circularity score is 0.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central result is a fitted deep network; all numbers are benchmark scores, not derived constants. The ledger entries are the external pretrained components and measurement choices the SOTA claim rests on.

free parameters (4)

End-to-end trained diffusion/garment UNet weights and FIA projection layers = learned on VITON-HD or DressCode training split
All reported metrics depend on these fitted weights; no pretrained checkpoint is released.
Training hyperparameters (learning rate, batch size, steps, resolution) = 1e-5, 64, 100k, 512x384
Chosen by hand in Section 4.1; no sensitivity analysis is given.
Frozen Flow Guider (DSDM) weights = D4-VTON pretrained flow estimator, not released
Produces the dense flow F used in every FIA module; its accuracy is a fixed input, not adapted.
FashionCLIP feature encoder weights = pretrained FashionCLIP, not released
Spatial embeddings S come from this fixed model; only a one-row OpenCLIP ablation is provided.

assumptions (5)

domain assumption Stable Diffusion v2.1 inpainting is a suitable base for virtual try-on generation.
The entire architecture is built on it (Section 3.4 and Implementation Details), and no comparison against other diffusion bases is made.
domain assumption The frozen DSDM flow maps used as F are sufficiently aligned with actual garment deformation.
Section 3.2 adopts DSDM without fine-tuning and without analyzing failure cases.
domain assumption FashionCLIP's highest-layer feature map retains spatial layout while encoding high-level garment semantics.
Section 3.3 states this property; only a limited OpenCLIP comparison is provided.
domain assumption FID/SSIM/PSNR/LPIPS computed from a single run and quoted from other papers are comparable across methods.
Section 4.1 acknowledges resampling sensitivity, yet Table 1 mixes bilinear and cubic downsampling conditions.
domain assumption DensePose, OpenPose, and human parsing are available and accurate for test and train images.
Section 3.4 uses these external tools to build the conditioning triplet.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Learning Implicit Features with Flow Infused Attention for Realistic Virtual Try-On." pith.science (2026). https://pith.science/paper/63RULTRN

@misc{pith2026241211435,
  author       = {Pith},
  title        = {Pith review of: Learning Implicit Features with Flow Infused Attention for Realistic Virtual Try-On},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/63RULTRN}},
  note         = {Machine review of arXiv:2412.11435}
}

read the original abstract

Image-based virtual try-on is challenging since the generated image should fit the garment to model images in various poses and keep the characteristics and details of the garment simultaneously. A popular research stream warps the garment image firstly to reduce the burden of the generation stage, which relies highly on the performance of the warping module. Other methods without explicit warping often lack sufficient guidance to fit the garment to the model images. In this paper, we propose FIA-VTON, which leverages the implicit warp feature by adopting a Flow Infused Attention module on virtual try-on. The dense warp flow map is projected as indirect guidance attention to enhance the feature map warping in the generation process implicitly, which is less sensitive to the warping estimation accuracy than an explicit warp of the garment image. To further enhance implicit warp guidance, we incorporate high-level spatial attention to complement the dense warp. Experimental results on the VTON-HD and DressCode dataset significantly outperform state-of-the-art methods, demonstrating that FIA-VTON is effective and robust for virtual try-on.

Figures

Figures reproduced from arXiv: 2412.11435 by the authors.

**Figure 2.** (a) Overview of FIA-VTON, illustrating the main components: a pair of VAE Encoder and Decoder, a Garment Net, and a [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Qualitative comparison on VITON-HD dataset [ [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: Qualitative comparison on the DressCode [ [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

**Figure 5.** Figure 5: Ablation study on our FIA-VTON. ing Fashion-CLIP with OpenCLIP, and (4) the full model. Effectiveness of the Flow Guider module. As shown in [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

**Figure 6.** Figure 6: Ablation study on Flow Infused Attention (FIA). [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

**Figure 7.** Figure 7: Qualitative comparison in the wild scenarios. Compared with state-of-the-art methods (OOTDiffusion, CAT-VTON). Our method generates more natural images that seamlessly combine background, person, and garment in complex scenarios. Zoom in for more details. ditions. As s…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

MS-DETR: Towards Effective Video Moment Retrieval and Highlight Detection by Joint Motion-Semantic Learning
cs.CV 2025-07 conditional novelty 6.0 of 10

MS-DETR improves moment retrieval and highlight detection by disentangling motion and semantic video features, sharing task information between the two tasks, and training on generated auxiliary captions.

Reference graph

Works this paper leans on

31 extracted references · 21 canonical work pages · cited by 1 Pith paper

[1]

Stable diffusion 2.1

Stability AI. Stable diffusion 2.1. https : / / huggingface . co / stabilityai / stable - diffusion-2-1, 2023. Accessed: 2023-11-15. 5

work page 2023
[2]

Demystifying mmd gans

Mikołaj Bi ´nkowski, Danica J Sutherland, Michael Arbel, and Arthur Gretton. Demystifying mmd gans. arXiv preprint arXiv:1801.01401, 2018. 5

arXiv 2018
[3]

Realtime multi-person 2d pose estimation using part affinity fields

Zhe Cao, Tomas Simon, Shih-En Wei, and Yaser Sheikh. Realtime multi-person 2d pose estimation using part affinity fields. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 7291–7299, 2017. 5

work page 2017
[4]

Viton-hd: High-resolution virtual try-on via misalignment-aware normalization

Seunghwan Choi, Sunghyun Park, Minsoo Lee, and Jaegul Choo. Viton-hd: High-resolution virtual try-on via misalignment-aware normalization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 14131–14140, 2021. 2, 5, 6

work page 2021
[5]

Improving diffusion models for vir- tual try-on

Yisol Choi, Sangkyung Kwak, Kyungmin Lee, Hyungwon Choi, and Jinwoo Shin. Improving diffusion models for vir- tual try-on. arXiv preprint arXiv:2403.05139, 2024. 5

arXiv 2024
[6]

Catvton: Concatenation is all you need for virtual try-on with diffusion models, 2024

Zheng Chong, Xiao Dong, Haoxiang Li, Shiyue Zhang, Wenqing Zhang, Xujie Zhang, Hanqing Zhao, and Xiaodan Liang. Catvton: Concatenation is all you need for virtual try-on with diffusion models, 2024. 2

work page 2024
[7]

Diffusion mod- els beat gans on image synthesis

Prafulla Dhariwal and Alexander Nichol. Diffusion mod- els beat gans on image synthesis. Advances in neural information processing systems, 34:8780–8794, 2021. 2

work page 2021
[8]

Exploring Warping-Guided Features via Adaptive Latent Diffusion Model for Virtual try-on

Bo Gao, Junchi Ren, Fei Shen, Mengwan Wei, and Zijun Huang. Exploring Warping-Guided Features via Adaptive Latent Diffusion Model for Virtual try-on . In 2024 IEEE International Conference on Multimedia and Expo (ICME), pages 1–6, Los Alamitos, CA, USA, 2024. IEEE Computer Society. 3

work page 2024

Show all 31 references

[9]

Parser-free virtual try-on via distilling ap- pearance flows

Yuying Ge, Yibing Song, Ruimao Zhang, Chongjian Ge, Wei Liu, and Ping Luo. Parser-free virtual try-on via distilling ap- pearance flows. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8485– 8493, 2021. 3

2021
[10]

Taming the power of diffusion models for high-quality virtual try-on with appearance flow

Junhong Gou, Siyu Sun, Jianfu Zhang, Jianlou Si, Chen Qian, and Liqing Zhang. Taming the power of diffusion models for high-quality virtual try-on with appearance flow. In Proceedings of the 31st ACM International Conference on Multimedia, pages 7599–7607, 2023. 2, 5

2023
[11]

Vec- tor quantized diffusion model for text-to-image synthesis

Shuyang Gu, Dong Chen, Jianmin Bao, Fang Wen, Bo Zhang, Dongdong Chen, Lu Yuan, and Baining Guo. Vec- tor quantized diffusion model for text-to-image synthesis. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10696–10706, 2022. 2

2022
[12]

Viton: An image-based virtual try-on network

Xintong Han, Zuxuan Wu, Zhe Wu, Ruichi Yu, and Larry S Davis. Viton: An image-based virtual try-on network. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 7543–7552, 2018. 2

2018
[13]

Clothflow: A flow-based model for clothed person generation

Xintong Han, Xiaojun Hu, Weilin Huang, and Matthew R Scott. Clothflow: A flow-based model for clothed person generation. In Proceedings of the IEEE/CVF international conference on computer vision, pages 10471–10480, 2019. 3

2019
[14]

Style-based global appearance flow for virtual try-on

Sen He, Yi-Zhe Song, and Tao Xiang. Style-based global appearance flow for virtual try-on. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 3470–3479, 2022. 3

2022
[15]

Classifier-free diffusion guidance

Jonathan Ho and Tim Salimans. Classifier-free diffusion guidance. arXiv preprint arXiv:2207.12598, 2022. 2

2022 arXiv
[16]

Denoising dif- fusion probabilistic models

Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising dif- fusion probabilistic models. Advances in neural information processing systems, 33:6840–6851, 2020. 2

2020
[17]

Stableviton: Learning semantic corre- spondence with latent diffusion model for virtual try-on

Jeongho Kim, Guojung Gu, Minho Park, Sunghyun Park, and Jaegul Choo. Stableviton: Learning semantic corre- spondence with latent diffusion model for virtual try-on. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8176–8185, 2024. 2, 5

2024
[18]

High-resolution virtual try-on with misalignment and occlusion-handled conditions

Sangyun Lee, Gyojung Gu, Sunghyun Park, Seunghwan Choi, and Jaegul Choo. High-resolution virtual try-on with misalignment and occlusion-handled conditions. In European Conference on Computer Vision, pages 204–219. Springer, 2022. 5

2022
[19]

Toward accurate and realistic outfits visualization with atten- tion to details

Kedan Li, Min Jin Chong, Jeffrey Zhang, and Jingen Liu. Toward accurate and realistic outfits visualization with atten- tion to details. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 15546– 15555, 2021. 3

2021
[20]

Self- correction for human parsing

Peike Li, Yunqiu Xu, Yunchao Wei, and Yi Yang. Self- correction for human parsing. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(6):3260–3271, 2020. 5

2020
[21]

Fixing weight decay regularization in adam

Ilya Loshchilov, Frank Hutter, et al. Fixing weight decay regularization in adam. arXiv preprint arXiv:1711.05101, 5,

arXiv
[22]

Dress code: High- resolution multi-category virtual try-on

Davide Morelli, Matteo Fincato, Marcella Cornia, Federico Landi, Fabio Cesari, and Rita Cucchiara. Dress code: High- resolution multi-category virtual try-on. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2231–2235, 2022. 2, 5, 7

2022
[23]

Ladi-vton: Latent diffusion textual-inversion enhanced virtual try-on

Davide Morelli, Alberto Baldrati, Giuseppe Cartella, Mar- cella Cornia, Marco Bertini, and Rita Cucchiara. Ladi-vton: Latent diffusion textual-inversion enhanced virtual try-on. In Proceedings of the 31st ACM International Conference on Multimedia, pages 8580–8589, 2023. 2, 5

2023
[24]

On aliased resizing and surprising subtleties in gan evaluation

Gaurav Parmar, Richard Zhang, and Jun-Yan Zhu. On aliased resizing and surprising subtleties in gan evaluation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 11410–11420, 2022. 5

2022
[25]

High-resolution image synthesis with latent diffusion models

Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj ¨orn Ommer. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10684–10695, 2022. 3

2022
[26]

Improving virtual try- on with garment-focused diffusion models

Siqi Wan, Yehao Li, Jingwen Chen, Yingwei Pan, Ting Yao, Yang Cao, and Tao Mei. Improving virtual try- on with garment-focused diffusion models. arXiv preprint arXiv:2409.08258, 2024. 2, 3, 5

2024 arXiv
[27]

Toward characteristic- preserving image-based virtual try-on network

Bochao Wang, Huabin Zheng, Xiaodan Liang, Yimin Chen, Liang Lin, and Meng Yang. Toward characteristic- preserving image-based virtual try-on network. In Proceedings of the European conference on computer vision (ECCV), pages 589–604, 2018. 3

2018
[28]

Gp- vton: Towards general purpose virtual try-on via collabora- tive local-flow global-parsing learning

Zhenyu Xie, Zaiyu Huang, Xin Dong, Fuwei Zhao, Haoye Dong, Xijin Zhang, Feida Zhu, and Xiaodan Liang. Gp- vton: Towards general purpose virtual try-on via collabora- tive local-flow global-parsing learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Patte...

2023
[29]

D4-vton: Dynamic semantics disentangling for differential diffusion based virtual try-on

Zhaotong Yang, Zicheng Jiang, Xinzhe Li, Huiyu Zhou, Junyu Dong, Huaidong Zhang, and Yong Du. D4-vton: Dynamic semantics disentangling for differential diffusion based virtual try-on. arXiv preprint arXiv:2407.15111, 2024. 2, 3, 4, 5

2024 arXiv
[30]

The unreasonable effectiveness of deep features as a perceptual metric

Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shecht- man, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 586–595, 2018. 5

2018
[31]

Tryondiffusion: A tale of two unets

Luyang Zhu, Dawei Yang, Tyler Zhu, Fitsum Reda, William Chan, Chitwan Saharia, Mohammad Norouzi, and Ira Kemelmacher-Shlizerman. Tryondiffusion: A tale of two unets. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 4606– 4615, 2023. 2

2023

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

Stable diffusion 2.1

Stability AI. Stable diffusion 2.1. https : / / huggingface . co / stabilityai / stable - diffusion-2-1, 2023. Accessed: 2023-11-15. 5

work page 2023

[2] [2]

Demystifying mmd gans

Mikołaj Bi ´nkowski, Danica J Sutherland, Michael Arbel, and Arthur Gretton. Demystifying mmd gans. arXiv preprint arXiv:1801.01401, 2018. 5

arXiv 2018

[3] [3]

Realtime multi-person 2d pose estimation using part affinity fields

Zhe Cao, Tomas Simon, Shih-En Wei, and Yaser Sheikh. Realtime multi-person 2d pose estimation using part affinity fields. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 7291–7299, 2017. 5

work page 2017

[4] [4]

Viton-hd: High-resolution virtual try-on via misalignment-aware normalization

Seunghwan Choi, Sunghyun Park, Minsoo Lee, and Jaegul Choo. Viton-hd: High-resolution virtual try-on via misalignment-aware normalization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 14131–14140, 2021. 2, 5, 6

work page 2021

[5] [5]

Improving diffusion models for vir- tual try-on

Yisol Choi, Sangkyung Kwak, Kyungmin Lee, Hyungwon Choi, and Jinwoo Shin. Improving diffusion models for vir- tual try-on. arXiv preprint arXiv:2403.05139, 2024. 5

arXiv 2024

[6] [6]

Catvton: Concatenation is all you need for virtual try-on with diffusion models, 2024

Zheng Chong, Xiao Dong, Haoxiang Li, Shiyue Zhang, Wenqing Zhang, Xujie Zhang, Hanqing Zhao, and Xiaodan Liang. Catvton: Concatenation is all you need for virtual try-on with diffusion models, 2024. 2

work page 2024

[7] [7]

Diffusion mod- els beat gans on image synthesis

Prafulla Dhariwal and Alexander Nichol. Diffusion mod- els beat gans on image synthesis. Advances in neural information processing systems, 34:8780–8794, 2021. 2

work page 2021

[8] [8]

Exploring Warping-Guided Features via Adaptive Latent Diffusion Model for Virtual try-on

Bo Gao, Junchi Ren, Fei Shen, Mengwan Wei, and Zijun Huang. Exploring Warping-Guided Features via Adaptive Latent Diffusion Model for Virtual try-on . In 2024 IEEE International Conference on Multimedia and Expo (ICME), pages 1–6, Los Alamitos, CA, USA, 2024. IEEE Computer Society. 3

work page 2024