Pith. sign in

REVIEW 4 major objections 4 minor 1 cited by

JCo-MVTON: Jointly Controllable Multi-Modal Diffusion Transformer for Mask-Free Virtual Try-on

T0 review · 4 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read JCo-MVTON, a mask-free multi-modal diffusion transformer for virtual try-on, reports state-of-the-art scores on VITON-HD and DressCode and beats commercial systems in the wild.

desk verdict Plausible mask-free VTON system with strong DressCode numbers, but the benchmark claims need an explicit train/test-overlap statement and an honest abstract. read the letter →

arxiv 2508.17614 v1 pith:52LI34SL submitted 2025-08-25 cs.CV

Aowen Wang , Wei Li , Hao Luo , Mengxing Ao , Chenyu Zhu , Xinyang Li , Fan Wang This is my paper · ORCID
classification cs.CV
keywords virtualtry-onmask-freediffusiontransformerMM-DiTmulti-modalconditioningsynthetictripletdatasetself-attentionfusiongarmenttransfer
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 claims that virtual try-on does not need a human-body segmentation mask at inference time if the model is built as a multi-modal diffusion transformer that attends jointly to the person image and the target garment. If true, this makes virtual try-on simpler to deploy in real-world shopping applications, where reliable body parsing is often unavailable. The paper argues that mask-based pipelines are brittle because mask or warping errors cascade, while mask-free models were held back mainly by a lack of aligned person-garment-result training triplets. Its fix is a bidirectional data-generation loop: a Try-Off model recovers garment images from person photos, a mask-based model creates reference try-on images, human annotators filter, and the model itself regenerates sharper references over three rounds. The resulting system reports best or near-best scores on VITON-HD and DressCode without any mask at inference, and beats commercial systems in human evaluations on in-the-wild images.

What carries the argument

The load-bearing mechanism is the conditional MM-DiT block. The Multi-Modal Diffusion Transformer (MM-DiT) is a transformer architecture that fuses text and image tokens in shared self-attention; JCo-MVTON turns it into a multi-condition try-on model by tokenizing the noisy image, text prompt, reference person image, and garment image, concatenating them into one sequence, and processing them through three parallel query-key-value projection branches inside self-attention. A binary attention mask zeroes out any attention between reference tokens and garment tokens, so each condition models itself and attends to the noise and text stream without interfering with the other condition. A 'joint positional encoding' gives the noise and reference identical coordinates and places the garment at horizontally appended coordinates, preserving the background geometry of the source person photo. On the data side, the load-bearing step is the synthetic triplet loop: Try-Off extracts garments from person photos, a mask-based model builds initial reference images, human annotators filter, IC-LoRA expands style coverage, and the model itself regenerates sharper references. This loop is what makes the mask-free model trainable at scale.

What would settle it

Take the released model and run it on a held-out set of real e-commerce person-garment pairs that were never used in training, with manually checked ground-truth try-on images; if the LPIPS or FID gap versus a mask-based method on that set is much larger than the gap on VITON-HD or DressCode, the claimed wild generalization would not hold. A cheaper check: inspect whether the Try-Off model's recovered garments contain warped textures or borrowed background pixels, since those defects would propagate into the final try-on outputs.

Watch

Extended reading notes

Core claim

JCo-MVTON's central claim is that jointly injecting the reference person image and the target garment image into the self-attention layers of an MM-DiT diffusion transformer is enough for high-quality mask-free try-on, provided the two condition branches are kept from attending to each other and share a positional encoding that aligns the noise latent with the reference image. The paper also claims that the scarcity of aligned person-garment-result triplets can be overcome by a bidirectional generation loop: a self-supervised Try-Off model produces garment images from person photos, a mask-based model produces reference images, and manual filtering plus iterative regeneration with the model itself yields about 120K high-quality triplets. On VITON-HD, the method reports the best unpaired FID (9.561) and KID (2.700), two distribution-similarity scores, among compared methods, together with strong paired scores; on DressCode it reports the best scores across upper, lower, and dress categories on most metrics. In a five-metric human evaluation on real-world images, it is claimed to outperform three commercial try-on systems.

Load-bearing premise

The whole approach rests on the assumption that the synthetic triplets, garments recovered by Try-Off, references produced by a mask-based model, and then human-filtered, are faithful enough to serve as ground truth for real try-on, so any systematic artifact in those generated images would be learned as correct behavior.

Editorial extensions

If this is right

  • Deploying virtual try-on in the wild no longer requires a segmentation or parsing step; only a person photo and a garment image are needed at inference, so the pipeline is shorter and less prone to cascading mask errors.
  • The bidirectional generation loop, Try-Off for garments, mask-based try-on for references, human filtering, then model-based regeneration, can be iterated to grow training data across garment types and styles; the paper reports three rounds suffice.
  • Because the attention mask is defined by condition labels, additional condition branches beyond person and garment can be added without retraining the base backbone.
  • The reported numbers imply that mask-free models can top benchmark tables on unpaired and most category-specific metrics, not merely match mask-based competitors.
  • On VITON-HD, the method reports the best unpaired FID and KID among compared methods, and on DressCode most category metrics are best, showing mask-free models can lead rather than merely match.

Reading between the lines

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

  • An extension not explored in the paper is to train a control model on only the original public paired data, skipping the synthetic loop; if the synthetic-trained model wins mainly through overlap with its own generators, the wild-generalization claim would need qualification.
  • Because the mutual-exclusion mask forbids reference and garment branches from attending to each other entirely, allowing a small learned amount of cross-attention in later layers could improve fine-grained alignment of folds, seams, and shadows, a cheap ablation the authors leave open.
  • The same architecture and bidirectional loop should be able to run the inverse try-off task, removing a garment from a person, since the Try-Off generator already exists inside the data pipeline; a public try-off benchmark would make that capability measurable.
  • Full-parameter fine-tuning of the duplicated QKV branch beats LoRA in the paper, but at notable training cost; comparing higher-rank adapters or selective layer unfreezing under a fixed compute budget is a natural next experiment.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper proposes JCo-MVTON, a mask-free virtual try-on method built on a Multi-Modal Diffusion Transformer (MM-DiT) backbone. The authors inject reference-person and garment-image latents into the self-attention layers of FLUX through separate QKV branches, use mutually exclusive attention masks to prevent cross-attention between conditional branches, and introduce a concatenation-based positional encoding scheme. To create training data, they design a two-stage pipeline: Stage I starts from VITON and DressCode plus crawled Internet images, recovers garment images with a 'Try-Off' model and produces reference images with mask-based FLUX-Fill; Stage II applies human filtering, IC-LoRA-based domain expansion, and iterative model-assisted triplet regeneration. Experiments on VITON-HD and DressCode report strong unpaired FID/KID results and strong DressCode metrics, and human evaluation against Kling, OutfitAnyone, and GPT-4o is claimed to show superiority in real-world scenarios. The paper includes ablations for the conditional branch and positional encoding, and provides a project page and code link.

Significance. If the quantitative claims hold, the paper would be a useful contribution to mask-free virtual try-on: the architectural changes are concrete and reproducible, the data pipeline addresses a real scarcity of triplet training data, and the method is evaluated on standard public benchmarks plus qualitative wild-image comparisons. Strengths include a clearly described MM-DiT conditioning mechanism with an attention mask and positional-encoding design, an explicit two-stage data-generation protocol, ablations on VITON-HD, and the release of code and a demo API. The unpaired VITON-HD FID/KID improvement and the DressCode improvements across all three garment categories are notable. However, the headline 'state-of-the-art / significantly outperforming' claim is not consistently supported by the paired VITON-HD numbers, and the evaluation protocol has open questions about train/test overlap and the methodology of the human study. These issues are load-bearing for the central claim that the method generalizes better than prior art and commercial systems.

major comments (4)
  1. [Abstract and Table 1] The abstract states that JCo-MVTON 'achieves state-of-the-art performance on public benchmarks including DressCode, significantly outperforming existing methods,' but on VITON-HD paired evaluation the method trails GP-VTON in SSIM (0.8601 vs. 0.8701) and LPIPS (0.0891 vs. 0.0585). The paper should qualify the claim to specify that the advantage is primarily in unpaired FID/KID and on DressCode, and should discuss why paired perceptual similarity is worse despite better distributional metrics.
  2. [Section 4.2.1 (I.a, I.b, I.c) and Section 4.2.2 (II.c)] The training corpus is built from the same benchmark families used for evaluation: Stage I starts from VITON and DressCode, and Stage II iteratively regenerates triplets with a model trained on that corpus. The paper never states whether the VITON-HD and DressCode test splits are disjoint from the synthetic training triplets, nor whether ground-truth images from those test sets appear in any form in the generated reference images R. This is a load-bearing protocol gap: without an explicit overlap check or a held-out external test set, the unpaired FID/KID gains in Table 1 and the DressCode FIDp reductions in Table 2 could reflect training/evaluation distributional overlap rather than generalization. The authors should report a near-duplicate analysis between the training corpus and both test splits, or evaluate on a truly external benchmark.
  3. [Section 2 (Human Evaluation)] The human evaluation against commercial systems is presented as evidence of real-world superiority, but the methods section lacks essential details: the number of participants, whether they were blind to model identity, the number and selection criteria of test images, the exact scoring scale, and any statistical significance test. Without these details, the radar-chart comparison in Figure 2 cannot be independently assessed, and the statement that JCo-MVTON 'markedly outperforms competitors on all assessed metrics' is not verifiable. The authors should either provide a complete evaluation protocol or downgrade the claim to a qualitative observation.
  4. [Section 4.2.1 (I.b and I.c) and Section 4.2.2 (II.c)] The synthetic triplet generation assumes that the Try-Off model recovers accurate garment images, that FLUX-Fill produces correct try-on references, and that manual filtering removes systematic artifacts. Since the model is trained on these synthetic triplets, any systematic bias in garment texture, lighting, or folds could be learned as ground truth. The paper should provide a quantitative analysis of the synthetic data quality, e.g., human agreement rates in the filtering step, comparison of Try-Off outputs against real garment crops, and evidence that iterative regeneration does not amplify artifacts.
minor comments (4)
  1. [Throughout] There are several typos and inconsistent terms: 'Vitual' in the Section 4 title, 'Jonit' in Section 4.4, 'Ominicontrol' for OmniControl in Section 3.1, and inconsistent use of 'IC-LoRA' vs 'IC-LoRa' and 'MMDiT' vs 'MM-DiT'. These should be fixed in a revision.
  2. [Table 1 and Table 2] No error bars, confidence intervals, or significance tests are reported for any metric. Since some comparisons are close (e.g., DressCode upper SSIM 0.9123 vs. IDM-VTON 0.9125), the authors should report standard deviations over multiple evaluation runs or at least multiple random seeds.
  3. [Section 2.1] The bullet for the Dress scenario contains a formatting error: '(d) Dress: (c) Lower: JCo-MVTON tops all metrics; attains the best performance on all metrics except a slight lag behind Kling in Detail Consistency.' This should be rewritten for clarity.
  4. [Section 5, Implementation details and Datasets] The paper says 'tens of millions of e-commerce data' were collected, but the final training set is 141,734 triplets. The manuscript should clarify the relationship between these numbers, e.g., the filtering ratio and whether the e-commerce data overlaps with the VITON/DressCode-derived synthetic corpus.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: benchmark claims are externally anchored and the synthetic-data loop is a data-generation procedure, not a derivation from the prediction.

full rationale

The central claims are state-of-the-art quantitative results on VITON-HD and DressCode and superiority over commercial systems in human evaluation. Neither claim is obtained by fitting a parameter and then relabeling it as a prediction; the benchmark numbers compare generated images against public test-set ground truth or distribution statistics, and the human evaluation is an external comparison. The two-stage data pipeline (Try-Off garment recovery, FLUX-Fill reference generation, and iterative JCo-MVTON regeneration) is a bootstrapping data-augmentation scheme: the model is trained on synthetic triplets whose R images are produced by earlier generative models, but the final evaluation is not on those synthetic R images. The paper makes no load-bearing appeal to a uniqueness theorem or to prior work by the current authors; citations such as Try-Off, IC-LoRA, FLUX, and CatVTON are external and provide concrete architectural or data-generation components. The closest concern is that the synthetic corpus is seeded from VITON and DressCode and the paper does not state whether benchmark test splits are disjoint from the generated training pool; that is a potential evaluation-protocol or contamination risk, not a circularity in which a claimed prediction reduces to an input by construction. The human evaluation similarly lacks methodological details such as participant count, blinding, and statistical tests, but this weakens the evidence rather than making the claim circular. No equation in the paper defines the predicted output in terms of the reported metric, and no fitted value is renamed as a result. Accordingly, no circular step can be quoted and exhibited, and the appropriate finding is no significant circularity.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

Three hand-chosen quantities are listed. They are not fitted to the benchmark in a hidden way, but they shape the method and are not justified by a principled rule. The method leans on several external pretrained models and datasets (FLUX, BiRefNet, IC-LoRA, TryOffDiff, VITON/DressCode) and on the assumption that synthetic targets produced by mask-based try-on are valid ground truth. invented_entities: none; the paper introduces no new physical or conceptual entities beyond the model itself.

free parameters (3)
  • Number of iterative refinement rounds = 3
    The pipeline repeats stages II.a-II.c three times until performance 'saturates' (Section 4.2.2). This is a hand-chosen stopping criterion.
  • Training triplet count = 141,734
    The final training set size after manual filtering (Section 5, Datasets). It is a data-availability choice, not fitted to an independent objective.
  • Initial resolution and batch size = 512x384, batch 16; then 1024x768, batch 4
    Coarse-to-fine training schedule (Section 4.5). These hyperparameters are chosen by hand.
assumptions (6)
  • domain assumption The FLUX.1-dev pretrained diffusion transformer provides a strong generative prior that can be fine-tuned for virtual try-on.
    Stated in Section 4.3; the entire method wraps FLUX and inherits its strengths and limitations.
  • domain assumption The Try-Off model (architecture identical to JCo-MVTON) can recover clean garment images from a person image in a mask-free manner.
    Section 4.2.1 (I.b) relies on this for garment extraction; poorly recovered garments would corrupt the training triplets.
  • domain assumption BiRefNet accurately extracts garment foregrounds from the paired data.
    Section 4.2.1 (I.b) uses BiRefNet to build Try-Off training pairs; inaccurate segmentation propagates into the pipeline.
  • domain assumption The mask-based FLUX-Fill model produces correct 'reference' images (person wearing the target garment) that can serve as ground truth for training a mask-free model.
    Section 4.2.1 (I.c) generates R with mask-based try-on; if the mask-based outputs contain artifacts, the mask-free model learns those artifacts as targets.
  • domain assumption VITON-HD and DressCode are appropriate external benchmarks and their provided splits are used without contamination from the self-generated training data.
    Section 5 describes evaluation on these datasets, but Section 4.2.1 says the training set starts from VITON and DressCode; the paper does not explicitly guarantee disjoint splits.
  • standard math Standard transformer attention and softmax mathematics are valid.
    Used in Eq. (6)-(10); these are established results.

how reviews work

0 comments
Cite this review

Pith. "Pith review of JCo-MVTON: Jointly Controllable Multi-Modal Diffusion Transformer for Mask-Free Virtual Try-on." pith.science (2026). https://pith.science/paper/52LI34SL

@misc{pith2026250817614,
  author       = {Pith},
  title        = {Pith review of: JCo-MVTON: Jointly Controllable Multi-Modal Diffusion Transformer for Mask-Free Virtual Try-on},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/52LI34SL}},
  note         = {Machine review of arXiv:2508.17614}
}
read the original abstract

Virtual try-on systems have long been hindered by heavy reliance on human body masks, limited fine-grained control over garment attributes, and poor generalization to real-world, in-the-wild scenarios. In this paper, we propose JCo-MVTON (Jointly Controllable Multi-Modal Diffusion Transformer for Mask-Free Virtual Try-On), a novel framework that overcomes these limitations by integrating diffusion-based image generation with multi-modal conditional fusion. Built upon a Multi-Modal Diffusion Transformer (MM-DiT) backbone, our approach directly incorporates diverse control signals -- such as the reference person image and the target garment image -- into the denoising process through dedicated conditional pathways that fuse features within the self-attention layers. This fusion is further enhanced with refined positional encodings and attention masks, enabling precise spatial alignment and improved garment-person integration. To address data scarcity and quality, we introduce a bidirectional generation strategy for dataset construction: one pipeline uses a mask-based model to generate realistic reference images, while a symmetric ``Try-Off'' model, trained in a self-supervised manner, recovers the corresponding garment images. The synthesized dataset undergoes rigorous manual curation, allowing iterative improvement in visual fidelity and diversity. Experiments demonstrate that JCo-MVTON achieves state-of-the-art performance on public benchmarks including DressCode, significantly outperforming existing methods in both quantitative metrics and human evaluations. Moreover, it shows strong generalization in real-world applications, surpassing commercial systems.

Review history (2 revisions) →

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

  1. Boogu-Image-0.1: Boosting Open Agentic Multimodal Generation via Understanding under a Minimal Budget

    cs.CV 2026-07 conditional novelty 4.0 of 10

    An open-source image-generation family shows that agentic prompt rewriting and a stronger text encoder can lift quality to near closed-source levels with only 208.62M images and about $400K of training compute.

Reference graph

Works this paper leans on

49 extracted references · 11 canonical work pages · cited by 1 Pith paper

  1. [1]

    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 format.date year duplicate empty "emp...

  2. [2]

    EfficientVITON: An Efficient Virtual Try-On Model using Optimized Diffusion Process

    Mostafa Atef, Mariam Ayman, Ahmed Rashed, Ashrakat Saeed, Abdelrahman Saeed, and Ahmed Fares. Efficientviton: An efficient virtual try-on model using optimized diffusion process, 2025. https://arxiv.org/abs/2501.11776

    work page Pith review arXiv 2025

  3. [3]

    Sutherland, Michael Arbel, and Arthur Gretton

    Mikołaj Bińkowski, Danica J. Sutherland, Michael Arbel, and Arthur Gretton. Demystifying mmd gans, 2021. https://arxiv.org/abs/1801.01401

    arXiv 2021

  4. [4]

    Pemf-vto: Point-enhanced video virtual try-on via mask-free paradigm, 2025

    Tianyu Chang, Xiaohao Chen, Zhichao Wei, Xuanpu Zhang, Qing-Guo Chen, Weihua Luo, Peipei Song, and Xun Yang. Pemf-vto: Point-enhanced video virtual try-on via mask-free paradigm, 2025. https://arxiv.org/abs/2412.03021

    arXiv 2025

  5. [5]

    Diffusion-based approaches for virtual try-on: A comprehensive study

    Lei Chen, Ming Zhao, and Yi Sun. Diffusion-based approaches for virtual try-on: A comprehensive study. International Journal of Computer Vision, 132 0 (5): 0 1123--1140, 2024 a

    work page 2024

  6. [6]

    Wear-any-way: Manipulable virtual try-on via sparse correspondence alignment, 2024 b

    Mengting Chen, Xi Chen, Zhonghua Zhai, Chen Ju, Xuewen Hong, Jinsong Lan, and Shuai Xiao. Wear-any-way: Manipulable virtual try-on via sparse correspondence alignment, 2024 b . https://arxiv.org/abs/2403.12965

    arXiv 2024

  7. [7]

    Viton-hd: High-resolution virtual try-on via misalignment-aware normalization, 2021

    Seunghwan Choi, Sunghyun Park, Minsoo Lee, and Jaegul Choo. Viton-hd: High-resolution virtual try-on via misalignment-aware normalization, 2021. https://arxiv.org/abs/2103.16874

    arXiv 2021

  8. [8]

    Improving diffusion models for authentic virtual try-on in the wild, 2024

    Yisol Choi, Sangkyung Kwak, Kyungmin Lee, Hyungwon Choi, and Jinwoo Shin. Improving diffusion models for authentic virtual try-on in the wild, 2024. https://arxiv.org/abs/2403.05139

    arXiv 2024

Show all 49 references
  1. [9]

    Catvton: Concatenation is all you need for virtual try-on with diffusion models, 2025

    Zheng Chong, Xiao Dong, Haoxiang Li, Shiyue Zhang, Wenqing Zhang, Xujie Zhang, Hanqing Zhao, Dongmei Jiang, and Xiaodan Liang. Catvton: Concatenation is all you need for virtual try-on with diffusion models, 2025. https://arxiv.org/abs/2407.15886

    2025 arXiv

  2. [10]

    Scaling rectified flow transformers for high-resolution image synthesis, 2024

    Patrick Esser, Sumith Kulal, Andreas Blattmann, Rahim Entezari, Jonas Müller, Harry Saini, Yam Levi, Dominik Lorenz, Axel Sauer, Frederic Boesel, Dustin Podell, Tim Dockhorn, Zion English, Kyle Lacey, Alex Goodwin, Yannik Marek, and Robin Rombach. Scaling rectified flow transf...

    2024 arXiv

  3. [11]

    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, MM ’23, page 7599–7607. ACM, October ...

    2023

  4. [12]

    Any2anytryon: Leveraging adaptive position embeddings for versatile virtual clothing tasks, 2025

    Hailong Guo, Bohan Zeng, Yiren Song, Wentao Zhang, Chuang Zhang, and Jiaming Liu. Any2anytryon: Leveraging adaptive position embeddings for versatile virtual clothing tasks, 2025. https://arxiv.org/abs/2501.15891

    2025 arXiv

  5. [13]

    Viton: An image-based virtual try-on network

    Xingjie Han, Zhe Wu, and Yunhong Zhang. Viton: An image-based virtual try-on network. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2018 a

    2018

  6. [14]

    Xintong Han, Zuxuan Wu, Zhe Wu, Ruichi Yu, and Larry S. Davis. Viton: An image-based virtual try-on network, 2018 b . https://arxiv.org/abs/1711.08447

    2018 arXiv

  7. [15]

    Fleet, Mohammad Norouzi, and Tim Salimans

    Jonathan Ho, Chitwan Saharia, William Chan, David J. Fleet, Mohammad Norouzi, and Tim Salimans. Cascaded diffusion models for high fidelity image generation, 2021. https://arxiv.org/abs/2106.15282

    2021 arXiv

  8. [16]

    Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen

    Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. Lora: Low-rank adaptation of large language models, 2021. https://arxiv.org/abs/2106.09685

    2021 arXiv

  9. [17]

    In-context lora for diffusion transformers

    Lianghua Huang, Wei Wang, Zhi-Fan Wu, Yupeng Shi, Huanzhang Dou, Chen Liang, Yutong Feng, Yu Liu, and Jingren Zhou. In-context lora for diffusion transformers. arXiv preprint arXiv:2410.23775, 2024

    2024 arXiv

  10. [18]

    Enhancing e-commerce with virtual try-on technologies

    Farah Islam and Rohan Patel. Enhancing e-commerce with virtual try-on technologies. Journal of Retail Innovation, 7 0 (2): 0 45--58, 2023

    2023

  11. [19]

    Fitdit: Advancing the authentic garment details for high-fidelity virtual try-on, 2024

    Boyuan Jiang, Xiaobin Hu, Donghao Luo, Qingdong He, Chengming Xu, Jinlong Peng, Jiangning Zhang, Chengjie Wang, Yunsheng Wu, and Yanwei Fu. Fitdit: Advancing the authentic garment details for high-fidelity virtual try-on, 2024. https://arxiv.org/abs/2411.10499

    2024 arXiv

  12. [20]

    Progressive growing of gans for improved quality, stability, and variation, 2018

    Tero Karras, Timo Aila, Samuli Laine, and Jaakko Lehtinen. Progressive growing of gans for improved quality, stability, and variation, 2018. https://arxiv.org/abs/1710.10196

    2018 arXiv

  13. [21]

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

    2024

  14. [22]

    A survey of diffusion models in generative virtual try-on

    Sangho Lee and Jisoo Kim. A survey of diffusion models in generative virtual try-on. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45 0 (11): 0 3201--3217, 2023

    2023

  15. [23]

    Mask-free virtual try-on with implicit warping

    Xiao Liu, Antonio Perez, and Miguel Torres. Mask-free virtual try-on with implicit warping. In Winter Conference on Applications of Computer Vision (WACV), 2024

    2024

  16. [24]

    Flow straight and fast: Learning to generate and transfer data with rectified flow, 2022

    Xingchao Liu, Chengyue Gong, and Qiang Liu. Flow straight and fast: Learning to generate and transfer data with rectified flow, 2022. https://arxiv.org/abs/2209.03003

    2022 arXiv

  17. [25]

    Prodigy: An expeditiously adaptive parameter-free learner, 2024

    Konstantin Mishchenko and Aaron Defazio. Prodigy: An expeditiously adaptive parameter-free learner, 2024. https://arxiv.org/abs/2306.06101

    2024 arXiv

  18. [26]

    Dress code: High-resolution multi-category virtual try-on, 2022

    Davide Morelli, Matteo Fincato, Marcella Cornia, Federico Landi, Fabio Cesari, and Rita Cucchiara. Dress code: High-resolution multi-category virtual try-on, 2022. https://arxiv.org/abs/2204.08532

    2022 arXiv

  19. [27]

    Ladi-vton: Latent diffusion textual-inversion enhanced virtual try-on, 2023

    Davide Morelli, Alberto Baldrati, Giuseppe Cartella, Marcella Cornia, Marco Bertini, and Rita Cucchiara. Ladi-vton: Latent diffusion textual-inversion enhanced virtual try-on, 2023. https://arxiv.org/abs/2305.13501

    2023 arXiv

  20. [28]

    Pfdm: Pseudo-image guided diffusion models for mask-free virtual try-on

    Xiaoyu Niu, Qiang Sun, and Jun Liu. Pfdm: Pseudo-image guided diffusion models for mask-free virtual try-on. In European Conference on Computer Vision (ECCV), 2024

    2024

  21. [29]

    Scalable diffusion models with transformers, 2023

    William Peebles and Saining Xie. Scalable diffusion models with transformers, 2023. https://arxiv.org/abs/2212.09748

    2023 arXiv

  22. [30]

    pytorch-fid: FID Score for PyTorch

    Maximilian Seitzer. pytorch-fid: FID Score for PyTorch . https://github.com/mseitzer/pytorch-fid, August 2020. Version 0.3.0

    2020

  23. [31]

    Roformer: Enhanced transformer with rotary position embedding, 2023

    Jianlin Su, Yu Lu, Shengfeng Pan, Ahmed Murtadha, Bo Wen, and Yunfeng Liu. Roformer: Enhanced transformer with rotary position embedding, 2023. https://arxiv.org/abs/2104.09864

    2023 arXiv

  24. [32]

    Ds-vton: High-quality virtual try-on via disentangled dual-scale generation, 2025

    Xianbing Sun, Yan Hong, Jiahui Zhan, Jun Lan, Huijia Zhu, Weiqiang Wang, Liqing Zhang, and Jianfu Zhang. Ds-vton: High-quality virtual try-on via disentangled dual-scale generation, 2025. https://arxiv.org/abs/2506.00908

    2025

  25. [33]

    Ominicontrol2: Efficient conditioning for diffusion transformers, 2025

    Zhenxiong Tan, Qiaochu Xue, Xingyi Yang, Songhua Liu, and Xinchao Wang. Ominicontrol2: Efficient conditioning for diffusion transformers, 2025. https://arxiv.org/abs/2503.08280

    2025 arXiv

  26. [34]

    Gomez, Lukasz Kaiser, and Illia Polosukhin

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. Attention is all you need, 2023. https://arxiv.org/abs/1706.03762

    2023 arXiv

  27. [35]

    Tryoffdiff: Virtual-try-off via high-fidelity garment reconstruction using diffusion models, 2024

    Riza Velioglu, Petra Bevandic, Robin Chan, and Barbara Hammer. Tryoffdiff: Virtual-try-off via high-fidelity garment reconstruction using diffusion models, 2024. https://arxiv.org/abs/2411.18350

    2024 arXiv

  28. [36]

    Mf-viton: High-fidelity mask-free virtual try-on with minimal input, 2025

    Zhenchen Wan, Yanwu xu, Dongting Hu, Weilun Cheng, Tianxi Chen, Zhaoqing Wang, Feng Liu, Tongliang Liu, and Mingming Gong. Mf-viton: High-fidelity mask-free virtual try-on with minimal input, 2025. https://arxiv.org/abs/2503.08650

    2025 arXiv

  29. [37]

    Unicombine: Unified multi-conditional combination with diffusion transformer, 2025 a

    Haoxuan Wang, Jinlong Peng, Qingdong He, Hao Yang, Ying Jin, Jiafu Wu, Xiaobin Hu, Yanjie Pan, Zhenye Gan, Mingmin Chi, Bo Peng, and Yabiao Wang. Unicombine: Unified multi-conditional combination with diffusion transformer, 2025 a . https://arxiv.org/abs/2503.09277

    2025 arXiv

  30. [38]

    Mv-vton: Multi-view virtual try-on with diffusion models, 2025 b

    Haoyu Wang, Zhilu Zhang, Donglin Di, Shiliang Zhang, and Wangmeng Zuo. Mv-vton: Multi-view virtual try-on with diffusion models, 2025 b . https://arxiv.org/abs/2404.17364

    2025 arXiv

  31. [39]

    Stablegarment: Garment-centric generation via stable diffusion, 2024 a

    Rui Wang, Hailong Guo, Jiaming Liu, Huaxia Li, Haibo Zhao, Xu Tang, Yao Hu, Hao Tang, and Peipei Li. Stablegarment: Garment-centric generation via stable diffusion, 2024 a . https://arxiv.org/abs/2403.10783

    2024 arXiv

  32. [40]

    Boow-vton: Body-oriented wearable virtual try-on with diffusion models

    Xi Wang, Yifan Li, and Lei Zhang. Boow-vton: Body-oriented wearable virtual try-on with diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024 b

    2024

  33. [41]

    Bovik, H.R

    Zhou 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, 13 0 (4): 0 600--612, 2004. doi:10.1109/TIP.2003.819861

    2004

  34. [42]

    Omnicontrol: Control any joint at any time for human motion generation, 2024

    Yiming Xie, Varun Jampani, Lei Zhong, Deqing Sun, and Huaizu Jiang. Omnicontrol: Control any joint at any time for human motion generation, 2024. https://arxiv.org/abs/2310.08580

    2024 arXiv

  35. [43]

    Gp-vton: Towards general purpose virtual try-on via collaborative local-flow global-parsing learning, 2023

    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 collaborative local-flow global-parsing learning, 2023. https://arxiv.org/abs/2303.13756

    2023 arXiv

  36. [44]

    Ootdiffusion: Outfitting fusion based latent diffusion for controllable virtual try-on, 2024

    Yuhao Xu, Tao Gu, Weifeng Chen, and Chengcai Chen. Ootdiffusion: Outfitting fusion based latent diffusion for controllable virtual try-on, 2024. https://arxiv.org/abs/2403.01779

    2024 arXiv

  37. [45]

    D ^4 -vton: Dynamic semantics disentangling for differential diffusion based virtual try-on, 2024

    Zhaotong Yang, Zicheng Jiang, Xinzhe Li, Huiyu Zhou, Junyu Dong, Huaidong Zhang, and Yong Du. D ^4 -vton: Dynamic semantics disentangling for differential diffusion based virtual try-on, 2024. https://arxiv.org/abs/2407.15111

    2024 arXiv

  38. [46]

    Efros, Eli Shechtman, and Oliver Wang

    Richard Zhang, Phillip Isola, Alexei A. Efros, Eli Shechtman, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric, 2018. https://arxiv.org/abs/1801.03924

    2018 arXiv

  39. [47]

    Boow-vton: Boosting in-the-wild virtual try-on via mask-free pseudo data training, 2024

    Xuanpu Zhang, Dan Song, Pengxin Zhan, Tianyu Chang, Jianhao Zeng, Qingguo Chen, Weihua Luo, and Anan Liu. Boow-vton: Boosting in-the-wild virtual try-on via mask-free pseudo data training, 2024. https://arxiv.org/abs/2408.06047

    2024 arXiv

  40. [48]

    Easycontrol: Adding efficient and flexible control for diffusion transformer, 2025

    Yuxuan Zhang, Yirui Yuan, Yiren Song, Haofan Wang, and Jiaming Liu. Easycontrol: Adding efficient and flexible control for diffusion transformer, 2025. https://arxiv.org/abs/2503.07027

    2025 arXiv

  41. [49]

    Bilateral reference for high-resolution dichotomous image segmentation

    Peng Zheng, Dehong Gao, Deng-Ping Fan, Li Liu, Jorma Laaksonen, Wanli Ouyang, and Nicu Sebe. Bilateral reference for high-resolution dichotomous image segmentation. arXiv preprint arXiv:2401.03407, 2024

    2024 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.