Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

SwiftTry: Fast and Consistent Video Virtual Try-On with Diffusion Models

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read The paper claims that reusing cached denoising features across shifted chunks makes video virtual try-on temporally consistent and about 60 percent faster than prior diffusion-based approaches.

desk verdict A plausible inference-time caching trick and a useful new dataset, but the evaluation numbers are internally inconsistent and the central speed-consistency claim is not supported by the paper's own tables. read the letter →

arxiv 2412.10178 v2 pith:P4OIPYCD submitted 2024-12-13 cs.CV cs.AI

classification cs.CVcs.AI
keywords videovirtualtry-ondiffusionmodelstemporalconsistencyfeaturecachinginpaintingShiftTikTokDressdatasetinferenceacceleration
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

SwiftTry claims that video virtual try-on—replacing the garment on a moving person while keeping the rest of the video intact—can be made both fast and temporally consistent by viewing it as conditional video inpainting and by avoiding the expensive overlapping-chunk stitching used by previous diffusion-based methods. Its inference scheme, ShiftCaching, processes non-overlapping chunks whose boundaries shift at each denoising step, and for half the frames it reuses cached features from earlier steps instead of recomputing the deep UNet blocks. A Masked Temporal Attention variant keeps those cached features from contaminating fresher ones. The paper also introduces TikTokDress, a high-resolution 817-video benchmark with complex backgrounds and movements. On VVT and TikTokDress, the method reports better video-consistency metrics than ViViD while running at 2.27 FPS, about 60 percent faster.

What carries the argument

ShiftCaching is the central mechanism. At each denoising step, the video is divided into non-overlapping chunks of N frames, and the chunk boundaries shift by a fixed offset of four frames between steps, so different frames meet at chunk edges over the sampling trajectory. To cut cost, only every other frame is fully propagated through the UNet; the other half is computed partially with features cached from the previous fully computed step (t+1 for one half, t+2 for the other). Masked Temporal Attention, specifically the Half Attention variant, then restricts the softmax so that lower-quality cached features can attend to high-quality fresh features while fresh features do not attend to stale ones, preventing errors from flowing backward through time while preserving temporal coherence.

What would settle it

Take a trained SwiftTry model and run it on a long clip with fast arm movements or a sudden lighting change exactly at a chunk boundary; if the cached-feature halves show visible texture smearing, flicker, or garment distortion that the fully computed frames do not, the cached-feature assumption is violated. A quantitative version would compare the Half Attention variant against full computation on such clips: if VFID_I3D on the cached halves is systematically worse than on fully computed frames, the temporal layers have not absorbed the distribution shift.

Watch

Extended reading notes

Core claim

On its own terms, the discovery is that the temporal-coherence problem in video try-on splits into two parts that can be solved separately: giving the diffusion model a temporal attention pathway so frames actually exchange information, and then feeding that pathway with features that are cheap but not stale. The authors show that by shifting chunk boundaries by four frames at each denoising step and caching the features of every other frame from the previous step, the model escapes the overlap-recompute trade-off documented in their Table 1. The remaining distribution shift—half the chunk's features come from timestep t+2 and half from t+1—is handled by a Half Attention mask that lets stale features read, but not write into, fresher features. With this machinery, the paper reports consistent long-video try-on at 2.27 FPS on an A100, outperforming baselines on VFID-based video-consistency metrics on both VVT and TikTokDress.

Load-bearing premise

The load-bearing premise is that temporal attention layers trained on fully computed features remain reliable when, at inference, half the frame features in a chunk are reused from earlier denoising steps that the network has never been trained on.

Editorial extensions

If this is right

  • Long video try-on no longer needs to recompute overlapping frames, so the cost of temporal consistency becomes roughly independent of overlap size, which is what pushes inference to 2.27 FPS.
  • Because temporal attention is the only module trained in the second stage and only it is masked at inference, the approach transfers the image try-on model's garment fidelity directly to video.
  • On the VVT benchmark, the method reports VFID_I3D of 3.589 and VFID_RN of 0.534, improving over ViViD's 6.788 and 0.853 while keeping LPIPS and SSIM competitive.
  • With TikTokDress, evaluation moves to higher-resolution, in-the-wild clips, so the reported gains are tied to harder backgrounds and movements than previous public benchmarks.
  • Matching training and inference chunk length matters: the paper finds that 16-frame chunks give the best results, so the temporal layers are specialized to that chunk size.

Reading between the lines

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

  • The same cached-feature trick should transfer to other video diffusion tasks that already have temporal attention, such as human animation or object editing; the only task-specific piece is the Half Attention mask's assumption that fresher features are always the more reliable ones.
  • A natural testable extension is to train temporal attention directly on mixed cached and fresh inputs; if that closes the distribution gap, the cache ratio could be raised beyond 50 percent for even larger speedups.
  • The supplement's mask-error analysis implies that a large share of try-on artifacts in current datasets comes from agnostic-mask mistakes rather than the generative model, so improving automatic segmentation may yield bigger quality gains than further architecture work.
  • The paper's fixed shift of four frames and chunk length of sixteen are both tuned hyperparameters; ablating the shift size on high-motion videos would show whether the consistency guarantee degrades gracefully when motion is fast.
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 SwiftTry, a diffusion-based video virtual try-on method built from a two-stage pipeline: an image try-on model in the style of StableVITON (Garment UNet, pose control, Reference Attention) and a video stage that adds temporal attention layers initialized from AnimateDiff. For efficient long-video inference, it introduces ShiftCaching, which shifts chunk boundaries across denoising steps and partially computes frames using cached features, together with a Masked Temporal Attention (Half Attention) variant designed to handle mismatched cached features. The paper also contributes TikTokDress, a dataset of 817 high-resolution short-form videos with manually refined garment masks, diverse backgrounds, and complex motions. Experiments on VVT and TikTokDress claim that SwiftTry outperforms baselines in video consistency and runs at 2.27 FPS, about 60 percent faster than the re-evaluated ViViD baseline.

Significance. If the claims hold, ShiftCaching would be a practical contribution to long-video try-on inference, and TikTokDress would address a real gap in dataset diversity and mask quality. The paper's concrete assets are a reproducible speed comparison against one re-run baseline, a documented annotation pipeline for TikTokDress, and a set of ablations over attention-mask designs and chunk lengths. However, the quantitative support for the central claim is currently not internally consistent: the VVT main table reports VFIDI3D = 3.589 while all ShiftCaching ablations on the same dataset report values around 8.7-9.3, and the TikTokDress comparison shows the proposed technique degrading VFIDI3D from 3.865 to 4.231. These discrepancies, combined with a mixed protocol for baseline numbers, prevent the paper from establishing its headline speed-consistency claim.

major comments (4)
  1. [Tables 3, 5, 6, 7] Table 3 reports SwiftTry VFIDI3D = 3.589 on the VVT dataset, but every SwiftTry variant in Tables 5, 6, and 7 on the same dataset reports VFIDI3D between 8.679 and 9.312. No stated difference - fixed vs random shift, partial-computation ratio, attention-mask type, or chunk length - accounts for this more than five-point gap. The main table and the ablations appear to have been produced under different evaluation protocols, or one set of numbers is erroneous. Because the abstract's claim of superior video consistency rests on Table 3, this inconsistency is load-bearing and must be resolved by reporting the exact protocol for each table or recomputing all numbers under one protocol.
  2. [Table 4] Table 4 directly compares SwiftTry without ShiftCaching (VFIDI3D = 3.865) with SwiftTry with ShiftCaching (VFIDI3D = 4.231) on TikTokDress, a relative degradation of roughly 9 percent in the temporal-consistency metric. The paper nonetheless claims that ShiftCaching 'maintains temporal consistency' (Abstract) and that it 'ensures temporal smoothness' (Conclusion). No error bars or repeated runs are provided, so it is unclear whether the degradation is statistically significant, but as presented the data contradict the stated claim. The authors should report variance across seeds or videos and either demonstrate that the difference is within noise or qualify the claim as a speed-accuracy trade-off.
  3. [Methods (ShiftCaching), Fig. 4, Table 6] The temporal attention layers are trained only on fully computed features at a consistent noise level within a chunk, but at inference ShiftCaching feeds half of each chunk with cached features from timesteps t+2 and t+1 (Fig. 3). The Masked Temporal Attention (Half Attention) is introduced to reconcile this mismatch, yet the model is never trained with mixed cached/current inputs or with the mask applied. The paper offers no analysis of the resulting distribution shift (for example, feature distances between cached and current features) and no evidence that the mask is sufficient on videos outside the test set. Given that Table 4 shows degraded VFIDI3D when ShiftCaching is enabled, this untested assumption is directly relevant to the central speed-consistency claim. The authors should address it, for instance by fine-tuning temporal layers with partial computation or by explicitly ablating the mismatch.
  4. [Table 3 and 'Comparisons with Prior Approaches'] Table 3 mixes a re-run of ViViD (marked with a dagger) with numbers reported by other papers under unspecified protocols, so the comparison is not controlled. The text further states that SwiftTry 'excels in the VFID metric', but Tunnel Try-On has a better (lower) VFIDI3D (3.345 vs 3.589) and better LPIPS (0.054 vs 0.066) in the same table. The claim of outperforming baselines in video consistency is therefore contradicted by the table itself, and the protocol mismatch prevents adjudicating the comparison. The authors should state the evaluation protocol for each baseline, use identical masks/pose inputs where feasible, and revise the claim to match the reported numbers.
minor comments (5)
  1. [ShiftCaching Technique] The shift value Delta is introduced only through the example in Fig. 3 with Delta = 4; the paper does not specify how Delta is chosen or whether it is tuned per dataset. Please define the shift schedule precisely.
  2. [Table 5] The 'Random Shift' variant in Table 5 is not defined in the main text; it is unclear how random shifts are sampled (per step, per chunk) and whether the random seed is fixed across runs.
  3. [All quantitative tables] No error bars, confidence intervals, or repeated-run statistics are reported for any metric. Given the small differences in Tables 4 and 6, this makes it impossible to assess the significance of the reported improvements or degradations.
  4. [Supplementary, Fig. 9 and main text] The main text says the mask 'set specific values in the attention matrix to 0', while the supplementary caption says white cells represent 0 and gray cells represent -infinity. Please align the mask semantics (0 vs -infinity before softmax) across the main text and supplement.
  5. [Reproducibility] The paper does not mention the release of code or checkpoints. Given the internal inconsistencies in the reported numbers, public release of the evaluation code and model weights would substantially strengthen the reproducibility of the claims.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: SwiftTry's speed and consistency claims rest on externally benchmarked experiments, not on equations that reduce to their own inputs; the reported VFID inconsistencies are a correctness risk, not circularity.

full rationale

Walking the derivation chain, the paper's architecture (image try-on stage plus temporal-attention stage) and ShiftCaching are evaluated against external baselines such as ViViD, StableVITON, FW-GAN, and ClothFormer on standard metrics including LPIPS, SSIM, VFID, and FPS. No parameter is fit to a target metric and then reported as a prediction of that same metric; the speed gain is a measured FPS number, and the consistency claim is supported or contradicted by VFIDI3D comparisons rather than by construction. The TikTokDress dataset is used for both training and testing, but this is standard supervised evaluation practice and does not make the result circular, since the method is not defined in terms of the dataset's labels or the evaluation metric. The citations to AnimateDiff, DeepCache, StableVITON, and AnimateAnyone are external prior work used for initialization or inspiration; there is no self-citation chain that forces the paper's conclusions, and the authors' own prior work is not load-bearing. The only significant concern is an internal inconsistency: Table 3 reports VFIDI3D = 3.589 on VVT while Tables 5 and 6 report values around 8.7 to 9.0 for the same dataset under the same method, and Table 4 shows ShiftCaching increasing VFIDI3D from 3.865 to 4.231 on TikTokDress. Those problems undermine the empirical support for the central claim, but they are issues of experimental reporting and statistical validity, not circular derivation. No equation in the paper is equivalent to its own input by construction, and no prediction is a renamed fitted value. Accordingly, the circularity score is 0.

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

The central claim rests on standard pretrained components (SD 1.5, AnimateDiff), a self-supervised training paradigm, and a hand-designed caching mechanism. No new physical entities or laws are introduced. The free parameters listed are design choices selected via ablations.

free parameters (4)
  • Chunk size N = 16
    Ablation in Table 7 shows N=16 beats 8 and 24; the model is trained with N=16, so this design choice is load-bearing for the speed/quality tradeoff.
  • Shift step delta = not explicitly stated in experiments (illustrated as 4 in Figure 3)
    The shifting mechanism's step size is a hand-chosen hyperparameter; no ablation is reported for delta.
  • Partial computation ratio = 50%
    ShiftCaching uses 50% partially computed frames per chunk to achieve the reported 2.27 FPS; the paper states '50% of the frames' in the ablation study.
  • Attention mask type in Masked Temporal Attention = Half Attention
    Table 6 shows Half Attention gives the best VFID; this choice is required for quality when using cached features.
assumptions (5)
  • domain assumption Stable Diffusion 1.5 pretrained weights provide a suitable initialization for the Main and Garment UNets.
    The method relies on the pretrained SD 1.5 model and AnimateDiff temporal attention as a starting point (Section: Overall Architecture).
  • domain assumption Self-supervised training with masked garment regions and a target garment image is sufficient to learn video try-on.
    Because paired before/after videos of the same person are unavailable, the model is trained to inpaint the masked region guided by the garment image (Section: Methods, Problem Statement).
  • ad hoc to paper Temporal attention layers trained only on fully computed features remain effective on partially computed features with cached values from earlier timesteps.
    ShiftCaching introduces a train/inference distribution shift; the paper's ablations test this only internally (Section: ShiftCaching Technique).
  • domain assumption The manually corrected garment masks in TikTokDress are accurate and representative of real-world conditions.
    The dataset construction relies on human refinement after SAM 2 segmentation (Section: TikTokDress Dataset).
  • domain assumption VFID, VFIDI3D, VFIDRN, LPIPS, and SSIM capture the relevant quality and temporal consistency dimensions.
    These are standard metrics used in the field; the paper adopts them without validating their correlation with human perception (Section: Experiments, Metrics).

how reviews work

0 comments
Cite this review

Pith. "Pith review of SwiftTry: Fast and Consistent Video Virtual Try-On with Diffusion Models." pith.science (2026). https://pith.science/paper/P4OIPYCD

@misc{pith2026241210178,
  author       = {Pith},
  title        = {Pith review of: SwiftTry: Fast and Consistent Video Virtual Try-On with Diffusion Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/P4OIPYCD}},
  note         = {Machine review of arXiv:2412.10178}
}
read the original abstract

Given an input video of a person and a new garment, the objective of this paper is to synthesize a new video where the person is wearing the specified garment while maintaining spatiotemporal consistency. Although significant advances have been made in image-based virtual try-on, extending these successes to video often leads to frame-to-frame inconsistencies. Some approaches have attempted to address this by increasing the overlap of frames across multiple video chunks, but this comes at a steep computational cost due to the repeated processing of the same frames, especially for long video sequences. To tackle these challenges, we reconceptualize video virtual try-on as a conditional video inpainting task, with garments serving as input conditions. Specifically, our approach enhances image diffusion models by incorporating temporal attention layers to improve temporal coherence. To reduce computational overhead, we propose ShiftCaching, a novel technique that maintains temporal consistency while minimizing redundant computations. Furthermore, we introduce the TikTokDress dataset, a new video try-on dataset featuring more complex backgrounds, challenging movements, and higher resolution compared to existing public datasets. Extensive experiments demonstrate that our approach outperforms current baselines, particularly in terms of video consistency and inference speed. The project page is available at https://swift-try.github.io/.

Figures

Figures reproduced from arXiv: 2412.10178 by the authors.

Figure 1
Figure 1. Results of our SwiftTry compared to those of [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of Stage 2 of our SwiftTry framework (Note that stage 1 is similar, except the input is a single image [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Illustration of fully and partially computed frames [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (11 more)
Figure 5
Figure 5. Figure 5: Example videos from the TikTokDress dataset [PITH_FULL_IMAGE:figures/full_fig_p004_5.png]
Figure 6
Figure 6. Figure 6: SAM 2 failures due to its sensitivity to prompts [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: Qualitative results of our method on the TikTokDress dataset. [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 8
Figure 8. Figure 8: Qualitative comparison with prior method on the VVT dataset. [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]
Figure 9
Figure 9. Figure 9: The attention mask used in different masked tem [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: Impact of Mask Quality on Try-On Results. The [PITH_FULL_IMAGE:figures/full_fig_p011_10.png]
Figure 11
Figure 11. Figure 11: Examples of T-pose the highest clarity, as illustrated in [PITH_FULL_IMAGE:figures/full_fig_p011_11.png]
Figure 13
Figure 13. Figure 13: In video b.mp4 (see Supplemental Videos), our method consistently preserves garment textures throughout the video, outperforming ViViD in maintaining texture sta￾bility over time. Similarly, in video d.mp4 (see Supplemen￾tal Videos), the”GAP” letters on the garment pr…
Figure 12
Figure 12. Figure 12: Examples from our dataset, which feature diverse poses and high-quality garments. [PITH_FULL_IMAGE:figures/full_fig_p012_12.png]
Figure 13
Figure 13. Figure 13: Qualitative comparison between our method and ViViD. Our model achieves robust results over long videos, pro [PITH_FULL_IMAGE:figures/full_fig_p012_13.png]
Figure 14
Figure 14. Figure 14: Additional examples of our video try-on results on the TikTokDress dataset. The first row illustrates how effectively [PITH_FULL_IMAGE:figures/full_fig_p013_14.png]

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. ChronoTailor: Harnessing Attention Guidance for Fine-Grained Video Virtual Try-On

    cs.CV 2025-06 conditional novelty 5.0 of 10

    ChronoTailor combines region-aware attention guidance, temporal feature fusion, and multi-scale garment-pose alignment to produce state-of-the-art video virtual try-on results, and contributes the StyleDress dataset.

Reference graph

Works this paper leans on

50 extracted references · 20 canonical work pages · cited by 1 Pith paper

  1. [1]

    Chen, C.-Y.; Lo, L.; Huang, P.-J.; Shuai, H.-H.; and Cheng, W.-H. 2021. Fashionmirror: Co-attention feature-remapping virtual try-on with sequential template poses. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 13809--13818

  2. [2]

    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

  3. [3]

    Choi, Y.; Kwak, S.; Lee, K.; Choi, H.; and Shin, J. 2024. Improving Diffusion Models for Virtual Try-on. arXiv preprint arXiv:2403.05139

  4. [4]

    Dong, H.; Liang, X.; Shen, X.; Wang, B.; Lai, H.; Zhu, J.; Hu, Z.; and Yin, J. 2019 a . Towards Multi-Pose Guided Virtual Try-On Network. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)

  5. [5]

    Dong, H.; Liang, X.; Shen, X.; Wu, B.; Chen, B.-C.; and Yin, J. 2019 b . Fw-gan: Flow-navigated warping gan for video virtual try-on. In Proceedings of the IEEE/CVF international conference on computer vision, 1161--1170

  6. [6]

    Fang, Z.; Zhai, W.; Su, A.; Song, H.; Zhu, K.; Wang, M.; Chen, Y.; Liu, Z.; Cao, Y.; and Zha, Z.-J. 2024. ViViD: Video Virtual Try-on using Diffusion Models. arXiv preprint arXiv:2405.11794

  7. [7]

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

  8. [8]

    J.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; and Bengio, Y

    Goodfellow, I. J.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; and Bengio, Y. 2014. Generative Adversarial Networks. arXiv:1406.2661

Show all 50 references
  1. [9]

    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

  2. [10]

    Guo, Y.; Yang, C.; Rao, A.; Wang, Y.; Qiao, Y.; Lin, D.; and Dai, B. 2023. Animatediff: Animate your personalized text-to-image diffusion models without specific tuning. arXiv preprint arXiv:2307.04725

  3. [11]

    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

  4. [12]

    Hang, T.; Gu, S.; Li, C.; Bao, J.; Chen, D.; Hu, H.; Geng, X.; and Guo, B. 2023. Efficient diffusion training via min-snr weighting strategy. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 7441--7451

  5. [13]

    He, S.; Song, Y.-Z.; and Xiang, T. 2022. Style-based global appearance flow for virtual try-on. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 3470--3479

  6. [14]

    H.; and Lin, L

    He, Z.; Chen, P.; Wang, G.; Li, G.; Torr, P. H.; and Lin, L. 2024. WildVidFit: Video Virtual Try-On in the Wild via Image-Based Controlled Diffusion Models. arXiv preprint arXiv:2407.10625

  7. [15]

    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

  8. [16]

    Jafarian, Y.; and Park, H. S. 2021. Learning High Fidelity Depths of Dressed Humans by Watching Social Media Dance Videos. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 12753--12762

  9. [17]

    Jiang, J.; Wang, T.; Yan, H.; and Liu, J. 2022. Clothformer: Taming video virtual try-on in all module. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 10799--10808

  10. [18]

    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

  11. [19]

    P.; and Ba, J

    Kingma, D. P.; and Ba, J. 2014. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980

  12. [20]

    Kuppa, G.; Jong, A.; Liu, X.; Liu, Z.; and Moh, T.-S. 2021. ShineOn: Illuminating design choices for practical video-based virtual clothing try-on. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 191--200

  13. [21]

    Lee, S.; Gu, G.; Park, S.; Choi, S.; and Choo, J. 2022. High-Resolution Virtual Try-On with Misalignment and Occlusion-Handled Conditions. arXiv preprint arXiv:2206.14180

  14. [22]

    J.; Zhang, J.; and Liu, J

    Li, K.; Chong, M. J.; Zhang, J.; and Liu, J. 2021. Toward accurate and realistic outfits visualization with attention to details. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 15546--15555

  15. [23]

    Li, P.; Xu, Y.; Wei, Y.; and Yang, Y. 2020. Self-correction for human parsing. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(6): 3260--3271

  16. [24]

    Ma, X.; Fang, G.; and Wang, X. 2024. Deepcache: Accelerating diffusion models for free. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 15762--15772

  17. [25]

    Morelli, D.; Fincato, M.; Cornia, M.; Landi, F.; Cesari, F.; and Cucchiara, R. 2022. Dress code: high-resolution multi-category virtual try-on. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2231--2235

  18. [26]

    Pumarola, A.; Goswami, V.; Vicente, F.; De la Torre, F.; and Moreno-Noguer, F. 2019. Unsupervised image-to-video clothing transfer. In Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops, 0--0

  19. [27]

    V.; Carion, N.; Wu, C.-Y.; Girshick, R.; Doll \'a r, P.; and Feichtenhofer, C

    Ravi, N.; Gabeur, V.; Hu, Y.-T.; Hu, R.; Ryali, C.; Ma, T.; Khedr, H.; R \"a dle, R.; Rolland, C.; Gustafson, L.; Mintun, E.; Pan, J.; Alwala, K. V.; Carion, N.; Wu, C.-Y.; Girshick, R.; Doll \'a r, P.; and Feichtenhofer, C. 2024. SAM 2: Segment Anything in Images and Videos. ...

  20. [28]

    Rogge, L.; Klose, F.; Stengel, M.; Eisemann, M.; and Magnor, M. 2014. Garment replacement in monocular video sequences. ACM Transactions on Graphics (TOG), 34(1): 1--10

  21. [29]

    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

  22. [30]

    Tseng, J.; Castellon, R.; and Liu, K. 2023. Edge: Editable dance generation from music. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 448--458

  23. [31]

    von Platen, P.; Patil, S.; Lozhkov, A.; Cuenca, P.; Lambert, N.; Rasul, K.; Davaadorj, M.; Nair, D.; Paul, S.; Berman, W.; Xu, Y.; Liu, S.; and Wolf, T. 2022. Diffusers: State-of-the-art diffusion models. https://github.com/huggingface/diffusers

  24. [32]

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

  25. [33]

    Wang, T.-C.; Liu, M.-Y.; Zhu, J.-Y.; Liu, G.; Tao, A.; Kautz, J.; and Catanzaro, B. 2018 b . Video-to-video synthesis. arXiv preprint arXiv:1808.06601

  26. [34]

    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

  27. [35]

    Z.; Ge, Y.; Wang, X.; Lei, S

    Wu, J. Z.; Ge, Y.; Wang, X.; Lei, S. W.; Gu, Y.; Shi, Y.; Hsu, W.; Shan, Y.; Qie, X.; and Shou, M. Z. 2023. Tune-a-video: One-shot tuning of image diffusion models for text-to-video generation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 7623--7633

  28. [36]

    Xie, Z.; Huang, Z.; Dong, X.; Zhao, F.; Dong, H.; Zhang, X.; Zhu, F.; and Liang, X. 2023. GP-VTON: Towards General Purpose Virtual Try-on via Collaborative Local-Flow Global-Parsing Learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,...

  29. [37]

    Xu, Y.; Gu, T.; Chen, W.; and Chen, C. 2024 a . OOTDiffusion: Outfitting Fusion based Latent Diffusion for Controllable Virtual Try-on. arXiv e-prints, arXiv--2403

  30. [38]

    Xu, Z.; Chen, M.; Wang, Z.; Xing, L.; Zhai, Z.; Sang, N.; Lan, J.; Xiao, S.; and Gao, C. 2024 b . Tunnel Try-on: Excavating Spatial-temporal Tunnels for High-quality Virtual Try-on in Videos. arXiv preprint arXiv:2404.17571

  31. [39]

    H.; Yan, H.; Liu, J.-W.; Zhang, C.; Feng, J.; and Shou, M

    Xu, Z.; Zhang, J.; Liew, J. H.; Yan, H.; Liu, J.-W.; Zhang, C.; Feng, J.; and Shou, M. Z. 2023. Magicanimate: Temporally consistent human image animation using diffusion model. arXiv preprint arXiv:2311.16498

  32. [40]

    Yang, H.; Zhang, R.; Guo, X.; Liu, W.; Zuo, W.; and Luo, P. 2020. Towards Photo-Realistic Virtual Try-On by Adaptively Generating-Preserving Image Content. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

  33. [41]

    Yang, Z.; Zeng, A.; Yuan, C.; and Li, Y. 2023. Effective whole-body pose estimation with two-stages distillation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 4210--4220

  34. [42]

    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

  35. [43]

    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

  36. [44]

    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

  37. [45]

    Zheng, J.; Zhao, F.; Xu, Y.; Dong, X.; and Liang, X. 2024. VITON-DiT: Learning In-the-Wild Video Try-On from Human Dance Videos via Diffusion Transformers. arXiv preprint arXiv:2405.18326

  38. [46]

    Zhong, X.; Wu, Z.; Tan, T.; Lin, G.; and Wu, Q. 2021. Mv-ton: Memory-based video virtual try-on network. In Proceedings of the 29th ACM International Conference on Multimedia, 908--916

  39. [47]

    Zhou, D.; Wang, W.; Yan, H.; Lv, W.; Zhu, Y.; and Feng, J. 2022. Magicvideo: Efficient video generation with latent diffusion models. arXiv preprint arXiv:2211.11018

  40. [48]

    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

  41. [49]

    , " * 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...

  42. [50]

    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 gl...

Pith tools

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