Vanast: Virtual Try-On with Human Image Animation via Synthetic Triplet Supervision

Hyunsoo Cha; Wonjung Woo; Byungjun Kim; Hanbyul Joo

cs.CVApr 6, 2026

Vanast: Virtual Try-On with Human Image Animation via Synthetic Triplet Supervision

Authors: Hyunsoo Cha, Wonjung Woo, Byungjun Kim, Hanbyul Joo

Organizations: Seoul National University

Abstract

We present Vanast, a unified framework that generates garment-transferred human animation videos directly from a single human image, garment images, and a pose guidance video. Conventional two-stage pipelines treat image-based virtual try-on and pose-driven animation as separate processes, which often results in identity drift, garment distortion, and front-back inconsistency. Our model addresses these issues by performing the entire process in a single unified step to achieve coherent synthesis. To enable this setting, we construct large-scale triplet supervision. Our data generation pipeline includes generating identity-preserving human images in alternative outfits that differ from garment catalog images, capturing full upper and lower garment triplets to overcome the single-garment-posed video pair limitation, and assembling diverse in-the-wild triplets without requiring garment catalog images. We further introduce a Dual Module architecture for video diffusion transformers to stabilize training, preserve pretrained generative quality, and improve garment accuracy, pose adherence, and identity preservation while supporting zero-shot garment interpolation. Together, these contributions allow Vanast to produce high-fidelity, identity-consistent animation across a wide range of garment types.

Figures & tables

Figure 2: Overview of Vanast Pipeline. Our Vanast framework generates virtual try-on human animation videos from a human image, garment images, and a pose video. By incorporating scalable human-image and garment-image generation pipelines, our method avoids dataset-specific constraints and trains effectively at scale. The Dual Modules architecture ensures that the three conditioning signals, human image $\mathbf{I}^{\mathbf{G^{\prime}}}$ , garment images $\mathbf{G}$ , and pose video $\mathbf{K}$ , are faithfully reflected in the resulting video.

Figure 3: Samples of Synthetic Triplet Datasets. We show samples of the datasets used for generation and training. The triplet construction contributes to enabling the model to preserve identity while accurately transferring garments and producing animation videos that follow the target pose.

Figure 4: Qualitative Comparisons (Subject-to-Image-based). We compare our results with baselines constructed by combining subject-to-image generation and animation models. Our method produces the most accurate pose following and garment transfer while preserving identity with high fidelity.

Figure 5: Qualitative Comparisons (Virtual Try-On-based). We compare our results with baselines formed by combining image virtual try-on models with animation models. Our method achieves the most accurate pose following and garment transfer while preserving identity with the highest fidelity.

		Internet Dataset							ViViD Dataset
Img.Gen.	Animation	L ${}_{1}\downarrow$	PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$	FID $\downarrow$	VFID ${}_{\text{I3D}}\downarrow$	VFID ${}_{\text{ResNeXt}}\downarrow$	L ${}_{1}\downarrow$	PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$	FID $\downarrow$	VFID ${}_{\text{I3D}}\downarrow$	VFID ${}_{\text{ResNeXt}}\downarrow$
VisualCloze	StableAnimator	0.2266	11.08	0.7210	0.4676	210.37	34.48	1.69	0.2835	8.80	0.6355	0.6061	190.83	49.04	7.21
VisualCloze	DisPose	0.2590	9.49	0.6527	0.5884	205.68	40.49	1.93	0.3043	8.25	0.5880	0.6871	214.00	55.82	11.71
MOSAIC	StableAnimator	0.1875	12.20	0.7128	0.4400	158.76	32.36	2.34	0.2382	10.02	0.6617	0.5583	158.47	44.46	2.55
MOSAIC	DisPose	0.1714	12.36	0.7133	0.4641	155.18	35.24	1.54	0.2008	10.67	0.6686	0.5619	155.98	50.90	4.91
UNO	StableAnimator	0.2025	11.79	0.7148	0.4381	162.15	31.73	3.12	0.2556	9.63	0.6622	0.5577	158.31	45.39	2.95

Table 1: Quantitative Comparison with the Combination of Subject-to-Image and Animation Models. We compare our model with a baseline that combines a subject-to-image model and an animation model. Our model achieves the best performance across all metrics. Bold text indicates the best score in each column.

		Internet Dataset							ViViD Dataset
Img.Gen.	Animation	L ${}_{1}\downarrow$	PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$	FID $\downarrow$	VFID ${}_{\text{I3D}}\downarrow$	VFID ${}_{\text{ResNeXt}}\downarrow$	L ${}_{1}\downarrow$	PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$	FID $\downarrow$	VFID ${}_{\text{I3D}}\downarrow$	VFID ${}_{\text{ResNeXt}}\downarrow$
OOTDiffusion	StableAnimator	0.1305	13.91	0.7412	0.3428	174.79	33.17	2.55	0.2431	9.99	0.6566	0.5528	168.46	43.39	2.66
OOTDiffusion	DisPose	0.1143	14.86	0.7551	0.3325	172.94	31.48	1.68	0.2101	10.72	0.6421	0.5927	169.92	45.13	8.32
CatVTON	StableAnimator	0.1242	14.56	0.7649	0.3273	132.09	26.43	0.90	0.2415	10.04	0.6772	0.5441	167.27	43.12	4.62
CatVTON	DisPose	0.0987	15.87	0.7779	0.3238	120.69	25.00	1.05	0.2083	10.66	0.6430	0.6018	155.30	43.82	9.62
OmniTry	StableAnimator	0.1227	14.53	0.7671	0.3178	121.04	25.66	1.18	0.2372	10.10	0.6726	0.5425	161.33	43.36	4.21

Table 2: Quantitative Comparison with the Combination of Image Virtual Try-On and Animation Models. We compare our model with a baseline that combines a image virtual try-on model and an animation model. Our model achieves the best performance across all metrics. Bold text indicates the best score in each column.

Figure 6: Result of Single Garment Transfer. We present virtual try-on with human image animation results generated from a single garment image.

Method	L ${{}_{1}\downarrow}$	PSNR ${\uparrow}$	SSIM ${\uparrow}$	LPIPS ${\downarrow}$	FID ${\downarrow}$	VFID ${}_{\text{I3D}}\downarrow$	VFID ${}_{\text{ResNeXt}}\downarrow$
Single Module	0.1162	14.28	0.6609	0.3974	108.84	39.64	1.76
Backbone-LoRA	0.1359	13.17	0.6314	0.4502	120.97	42.47	1.87
w/o SynthHuman	0.1163	14.62	0.6653	0.3943	110.76	38.89	1.93
Ours	0.1069	14.74	0.6657	0.3673	104.59	35.60	1.21

Table 3: Ablation Study. We conduct ablation study for each component of our model and dataset configuration. Bold text indicates the best score in each column.

Figure 7: Ablation Study. We present the ablation study results for the lower garment transfer. The red box in the “Single Module” result demonstrates vulnerability to pose conditions. Both “Backbone-LoRA” and “w/o SynthHuman” fail to achieve accurate garment transfer, as indicated in blue box. In contrast, our full model produces results most similar to the ground truth.

Figure 8: Result of Multiple Garment Transfer. We present zero-shot garment transfer results where both upper and lower garments are transferred simultaneously. The logos and fine details of the garments are well preserved and accurately reflected in the generated animation videos.

Figure 9: Result of In-the-wild Garment Transfer. We present garment transfer results using in-the-wild garment images provided by the TikTokDress [ 21 ] dataset.

Figure 10: Result of Garment Interpolation. Without requiring any additional finetuning, our Vanast model performs zero-shot transfer of interpolated garments by GTM. $\gamma$ denotes a scalar interpolation weight.

Virtual Try-On

Virtual try-on generates images or videos showing people wearing selected garments while preserving appearance, fit, pose, and identity.

See all 55 papers

OmniVTON++: Training-Free Universal Virtual Try-On with Principal Pose Guidance
Feb 16, 2026Zhaotong Yang, Yong Du, Shengfeng He +5Virtual Try-On
School of Computer Science and Technology, Ocean University of China, Qingdao, China · School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing, China · Sanya Oceanographic Institution, Ocean University of China, Sanya, China +2
UniVVT: A Unified End-to-End Framework for High-Fidelity Video Virtual Try-on
Aug 6, 2026Yushe Cao, Shikun Feng, Fei Shen +5Virtual Try-On Identity-Preserving Video Generation
Tsinghua University · Zhongguancun Academy · National University of Singapore +2
OmniTryOn: Video Try-On Anything at Once!
Jun 7, 2026Changliang Xia, Chengyou Jia, Minnan Luo +3Virtual Try-On Temporal Consistency in Video Generation
Xi’an Jiaotong University Xi’an, China

Feb 16, 2026J/K move · Enter open · S save

OmniVTON++: Training-Free Universal Virtual Try-On with Principal Pose Guidance

Zhaotong Yang, Yong Du, Shengfeng He, Yuhui Li+4

School of Computer Science and Technology, Ocean University of China, Qingdao, China · School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing, China · Sanya Oceanographic Institution, Ocean University of China, Sanya, China+2

Virtual Try-On

Open paper AskPDF arXiv

Image-based Virtual Try-On (VTON) concerns the synthesis of realistic person imagery through garment re-rendering under human pose and body constraints. In practice, however, existing approaches are typically optimized for specific data conditions, making their deployment reliant on retraining and limiting their generalization as a unified solution. We present OmniVTON++, a training-free VTON framework designed for universal applicability. It addresses the intertwined challenges of garment alignment, human structural coherence, and boundary continuity by coordinating Structured Garment Morphing for correspondence-driven garment adaptation, Principal Pose Guidance for step-wise structural regulation during diffusion sampling, and Continuous Boundary Stitching for boundary-aware refinement, forming a cohesive pipeline without task-specific retraining. Experimental results demonstrate that OmniVTON++ achieves state-of-the-art performance across diverse generalization settings, including cross-dataset and cross-garment-type evaluations, while reliably operating across scenarios and diffusion backbones within a single formulation. In addition to single-garment, single-human cases, the framework supports multi-garment, multi-human, and anime character virtual try-on, expanding the scope of virtual try-on applications. The code is available at https://github.com/Jerome-Young/OmniVTON-PlusPlus.

Abstract

Virtual Try-On

Explore similar work

OmniVTON++: Training-Free Universal Virtual Try-On with Principal Pose Guidance

UniVVT: A Unified End-to-End Framework for High-Fidelity Video Virtual Try-on

OmniTryOn: Video Try-On Anything at Once!