Gaussian Wardrobe: Compositional 3D Gaussian Avatars for Free-Form Virtual Try-On

Zhiyi Chen; Hsuan-I Ho; Tianjian Jiang; Jie Song; Manuel Kaufmann; Chen Guo

cs.CVMar 4, 2026

Gaussian Wardrobe: Compositional 3D Gaussian Avatars for Free-Form Virtual Try-On

Authors: Zhiyi Chen, Hsuan-I Ho, Tianjian Jiang, Jie Song, Manuel Kaufmann, Chen Guo

Organizations: Department of Computer Science, ETH Zürich · HKUST(GZ) · HKUST · ETH AI Center, ETH Zürich

Abstract

We introduce Gaussian Wardrobe, a novel framework to digitalize compositional 3D neural avatars from multi-view videos. Existing methods for 3D neural avatars typically treat the human body and clothing as an inseparable entity. However, this paradigm fails to capture the dynamics of complex free-form garments and limits the reuse of clothing across different individuals. To overcome these problems, we develop a novel, compositional 3D Gaussian representation to build avatars from multiple layers of free-form garments. The core of our method is decomposing neural avatars into bodies and layers of shape-agnostic neural garments. To achieve this, our framework learns to disentangle each garment layer from multi-view videos and canonicalizes it into a shape-independent space. In experiments, our method models photorealistic avatars with high-fidelity dynamics, achieving new state-of-the-art performance on novel pose synthesis benchmarks. In addition, we demonstrate that the learned compositional garments contribute to a versatile digital wardrobe, enabling a practical virtual try-on application where clothing can be freely transferred to new subjects. Project page: https://ait.ethz.ch/gaussianwardrobe

Figures & tables

Figure 2: Gaussian Wardrobe digitalizes compositional neural avatars from multi-view videos. Our pipeline consists of two major components: (left) a compositional Gaussian representation and (right) a framework for learning neural garments. We first reconstruct a mesh template

\mathcal{M}

from the first video frame and segment it into body

\mathcal{M}_{b}

, garment templates

\mathcal{M}_{\{u,\ell\}}

in the zero-shaped canonical space. During training, each layer learns a separate U-Net

\mathcal{F}

to predict the parameters of 3D Gaussian primitives

\mathbf{M}

from pose-conditioned positional maps

\mathbf{P}

. We composite the 3D Gaussians

\mathcal{G}

from all layers and render RGB images

\hat{I}

and segmentation masks

\hat{S}

to compute the training loss

\mathcal{L}

. The learned neural garments are shape-agnostic and can seamlessly transfer to other subjects for avatar virtual try-on.

Figure 3: Exemplar of avatar virtual try-on. Given a reconstructed avatar (left), we replace its lower garment with a new skirt,

\mathcal{M}_{\ell}^{\prime}

. The combined avatar can be animated to a novel pose

\theta^{\prime}

, which may introduce minor penetration artifacts (middle). We resolve these artifacts on-the-fly during rendering with our online penetration detection algorithm (right).

Method	PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$
4D-DRESS [ 62 ]
Animatable Gaussians [ 32 ]	27.75	0.9577	0.0531
LayGA [ 33 ]	27.58	0.9574	0.0543
Gaussian Wardrobe (Ours)	28.06	0.9579	0.0527
ActorsHQ [ 23 ]
Animatable Gaussians [ 32 ]	27.92	0.9411	0.0418

Table 1: Evaluation of novel pose synthesis on 4D-DRESS and ActorsHQ datasets. Best results are highlighted in bold . Our method consistently outperforms all baselines on all evaluation metrics ( cf . Fig. 4 ).

Figure 4: Qualitative comparisons of novel pose synthesis on 4D-Dress and ActorsHQ datasets. Our method can better model the dynamics of free-form garments, such as skirts (top) and vests (middle), and generate realistic renderings with sharper facial and garment details. In contrast, the baseline methods suffer from artifacts, such as blurry faces and semi-transparent clothing, and fail to reproduce fine details like wrinkles or pockets.

Method	mIoU $\uparrow$	Recall $\uparrow$	F1-Score $\uparrow$
w/o $\mathcal{L}_{sg}$	0.848	0.898	0.924
w/o $\mathcal{L}_{reg}$	0.879	0.922	0.940
w/o $\mathcal{L}_{pe}$	0.883	0.927	0.942
Ours Full	0.893	0.936	0.947

Table 2: Ablation study of the loss terms. We assess the effectiveness of each loss term from the full objective. We rendered segmentation masks of garment layers and compared them against the ground-truth masks ( cf . Fig. 5 ). The results show that the full model achieves the best performance.

Figure 5: Visualization on the effectiveness of loss terms. We rendered segmentation masks of garment layers to visualize the impact of each regularization term. The evaluation shows that our full model produces cleanest results, while the absence of $\mathcal{L}_{\text{pe}}$ , $\mathcal{L}_{\text{reg}}$ and $\mathcal{L}_{sg}$ leads to self-penetration or irregular segmentation.

Figure 6: Importance of segmentation loss in virtual try-on. We visualize the results of virtual try-on using the variants our method trained with and without $\mathcal{L}_{sg}$ . The garment’s appearance becomes corrupted due to the entanglement of body and clothing.

Figure 7: Virtual try-on application. Our method enables flexible 3D virtual try-on across different subjects. The resulting avatars can be animated to a novel pose.

Figure 8: Visualization of penetration-aware rendering. Penetration-aware rendering removes small artifacts between garment layers.

Method	PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$
Animatable Gaussians [ 32 ]	30.51	0.9491	0.0332
LayGA [ 33 ]	30.36	0.9482	0.0337
Gaussian Wardrobe (Ours)	30.86	0.9489	0.0331

Table 3: Quantitative evaluation of novel view synthesis on ActorsHQ dataset. We use a held-out camera view to evaluate the performance of novel view synthesis.

Method	4D-DRESS-00127-Inner			4D-DRESS-00185-Inner			4D-DRESS-00127-Outer
Method	SSIM $\uparrow$	PSNR $\uparrow$	LPIPS $\downarrow$	SSIM $\uparrow$	PSNR $\uparrow$	LPIPS $\downarrow$	SSIM $\uparrow$	PSNR $\uparrow$	LPIPS $\downarrow$
AG [ 32 ]	0.9579	28.73	0.0503	0.9612	28.78	0.0485	0.9539	25.62	0.0609
LayGA [ 33 ]	0.9579	28.63	0.0506	0.9607	28.53	0.0502	0.9533	25.49	0.0626
Ours	0.9585	28.98	0.0502	0.9615	29.15	0.0478	0.9535	25.95	0.0604
Method	ActorsHQ-Actor05			ActorsHQ-Actor08			ActorsHQ-Actor01
Method	SSIM $\uparrow$	PSNR $\uparrow$	LPIPS $\downarrow$	SSIM $\uparrow$	PSNR $\uparrow$	LPIPS $\downarrow$	SSIM $\uparrow$	PSNR $\uparrow$	LPIPS $\downarrow$

Table 4: Full quantitative evaluation of novel pose synthesis on the 4D-DRESS and ActorsHQ dataset.

Figure 9: Importance of diffused skinning fields. To deform the template mesh into the canonical pose, we evaluate different strategies for querying skinning weights in inverse LBS. Nearest-neighbor querying fails to preserve topology near the armpits, whereas our diffused skinning fields produce smoother and more accurate garment meshes.

Figure 10: More results on virtual try-on. Our method supports flexible 3D try-on across different subjects, and the generated avatars can be animated into novel poses.

Figure 11: Results on ActorsHQ. Our method produces photorealistic 360° rendering under novel views and poses.

Figure 12: Results on ActorsHQ. Our method produces photorealistic 360° rendering under novel views and poses.

3D Garment Reconstruction

See all 15 papers

Forwardrobe: Garment-Aware Gaussian Avatars from a Single Image
Jul 31, 2026Daisheng Jin, Shuyun Wang, Ying He3D Gaussian Splatting 3D Reconstruction
Nanyang Technological University · Singapore · The University of Queensland +1
DAMA: Disentangled Body-Anchored Gaussians for Controllable Multi-Layered Avatars
May 20, 2026Daniel Eskandar, Berna Kabadayi, Garvita Tiwari +13D Human Reconstruction 3D Reconstruction
University of Tübingen, Germany · Tübingen AI Center, Germany · Zuse School ELIZA, Germany +2
AGORA: Adversarial Generation Of Real-time Animatable 3D Gaussian Head Avatars
Dec 6, 2025Ramazan Fazylov, Sergey Zagoruyko, Aleksandr Parkin +23D Gaussian Splatting 3D Asset Generation
Mohamed bin Zayed University of Artificial Intelligence, Abu Dhabi, UAE · Polynome AI, Dubai, UAE · MTS AI, Moscow, Russia

Jul 31, 2026J/K move · Enter open · S save

Forwardrobe: Garment-Aware Gaussian Avatars from a Single Image

Daisheng Jin, Shuyun Wang, Ying He

Nanyang Technological University · Singapore · The University of Queensland+1

3D Gaussian Splatting 3D Reconstruction 3D Garment Reconstruction 3D Avatar Generation Human Avatar Animation 3D Human Reconstruction+2

Open paper AskPDF arXiv

Reconstructing animatable 3D human avatars from a single image remains particularly challenging for loose garments, whose geometry and motion cannot be adequately represented by body-aligned topology and skinning. We present Forwardrobe, a feed-forward framework for reconstructing garment-aware Gaussian avatars from a single image. Forwardrobe explicitly separates clothing from the body in canonical Gaussian space and equips the garment layer with continuity-aware geometry and skinning initialization, pose-conditioned non-rigid deformation, and appearance adaptation. These designs improve garment reconstruction and visual quality during animation, particularly for skirts and dresses. The separated garment layer additionally forms an independently controllable 3D asset, enabling garment editing, transfer, and 3D virtual try-on. Experiments demonstrate improved garment reconstruction quality and greater flexibility in garment manipulation compared with existing single-image avatar reconstruction methods.

Abstract

3D Garment Reconstruction

Explore similar work

Forwardrobe: Garment-Aware Gaussian Avatars from a Single Image

DAMA: Disentangled Body-Anchored Gaussians for Controllable Multi-Layered Avatars

AGORA: Adversarial Generation Of Real-time Animatable 3D Gaussian Head Avatars