Towards Photo-Realistic Virtual Try-On by Adaptively Generating↔Preserving Image Content

Yang, Han; Zhang, Ruimao; Guo, Xiaobao; Liu, Wei; Zuo, Wangmeng; Luo, Ping

doi:10.1109/cvpr42600.2020.00787

Cited by 235 publications

(267 citation statements)

References 32 publications

Supporting

Mentioning

267

Contrasting

Order By: Relevance

“…We perform visual comparison of our proposed method with CP-VTON [8] and ACGPN [10]. DP-VTON combines pixel transformation with feature transformation to generate accurate and clear warped clothing.…”

Section: Qualitative Resultsmentioning

confidence: 99%

“…On the other hand, we adopt Inception Score (IS) [17], Fréchet Inception Distance (FID) [18] and Peak Signal to Noise Ratio (PSNR) to evaluate the performance of our models on the quality of the generated try-on images. Table . 1 lists image quality (IS, FID and PSNR) and pair-wise structural similarity (SSIM) scores by CP-VTON [8], ACGPN [10] and DP-VTON. The quantitative metrics demonstrate the superiority of DP-VTON over other methods.…”

Section: Quantitative Resultsmentioning

confidence: 99%

“…It achieves the plausible try-on image synthesis, but the clothing details are mostly blurry. Most recently ACGPN [10] adaptively determines the generation or preservation of the distinct human parts in the synthesized image according to the predicted semantic layout. However, it is difficult to accurately generate the pixels in the parts of clothing and arms, especially when transferring a short-sleeve to a person in long-sleeve, or the opposite.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

DP-VTON: Toward Detail-Preserving Image-Based Virtual Try-on Network

Chang

Peng

et al. 2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

Image-based virtual try-on systems with the goal of transferring a target clothing item onto the corresponding region of a person have received great attention recently. However, it is still a challenge for the existing methods to generate photo-realistic try-on images while preserving non-target details(Fig. 1). To resolve this issue, we present a novel virtual try-on network, DP-VTON. First, a clothing warping module combines pixel transformation with feature transformation to transform the target clothing. Second, a semantic segmentation prediction module predicts a semantic segmentation map of the person wearing the target clothing. Third, an arm generation module generates arms of the reference image that will be changed after try-on. Finally, the warped clothing, semantic segmentation map, arms image and other non-target details (e.g. face, hair, bottom clothes) are fused together for try-on image synthesis. Extensive experiments demonstrate our system achieves the state-of-the-art virtual try-on performance both qualitatively and quantitatively. 1

show abstract

Section: Qualitative Resultsmentioning

confidence: 99%

Section: Quantitative Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

DP-VTON: Toward Detail-Preserving Image-Based Virtual Try-on Network

Chang

Peng

et al. 2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

show abstract

“…Our method outperforms the state of the art [Men et al 2020;Wang et al 2018a;Yang et al 2020] with respect to three components: body shape, photorealism, and skin color preservation on real images, as well as generalizing to other datasets, while only using unpaired data. While our method outperforms SOTA as it is, it can be further improved on real images to allow for complex texture patterns such as plaid, and specularities.…”

Section: Introductionmentioning

confidence: 82%

TryOnGAN

Lewis

Varadharajan

Kemelmacher-Shlizerman

2021

ACM Trans. Graph.

View full text Add to dashboard Cite

Given a pair of images---target person and garment on another person---we automatically generate the target person in the given garment. Previous methods mostly focused on texture transfer via paired data training, while overlooking body shape deformations, skin color, and seamless blending of garment with the person. This work focuses on those three components, while also not requiring paired data training. We designed a pose conditioned StyleGAN2 architecture with a clothing segmentation branch that is trained on images of people wearing garments. Once trained, we propose a new layered latent space interpolation method that allows us to preserve and synthesize skin color and target body shape while transferring the garment from a different person. We demonstrate results on high resolution 512 × 512 images, and extensively compare to state of the art in try-on on both latent space generated and real images.

show abstract

“…On a similar line, Jandial et al [16] presented a two-stage training pipeline consisting of a coarse-to-fine warping network and a texture transfer network conditioned on a learned segmentation mask and trained with a triplet loss strategy to further improve the quality of try-on results. More recently, Yang et al [17] proposed to generate the semantic layout of the target person and predict whether the corresponding image content needs to be generated or preserved, thus leading to more photo-realistic try-on results.…”

Section: Related Workmentioning

confidence: 99%

VITON-GT: An Image-based Virtual Try-On Model with Geometric Transformations

Fincato

Landi

Cornia

et al. 2021

2020 25th International Conference on Pattern Recognition (ICPR)

View full text Add to dashboard Cite

The large spread of online shopping has led computer vision researchers to develop different solutions for the fashion domain to potentially increase the online user experience and improve the efficiency of preparing fashion catalogs. Among them, image-based virtual try-on has recently attracted a lot of attention resulting in several architectures that can generate a new image of a person wearing an input try-on garment in a plausible and realistic way. In this paper, we present VITON-GT, a new model for virtual try-on that generates high-quality and photo-realistic images thanks to multiple geometric transformations. In particular, our model is composed of a two-stage geometric transformation module that performs two different projections on the input garment, and a transformation-guided try-on module that synthesizes the new image. We experimentally validate the proposed solution on the most common dataset for this task, containing mainly t-shirts, and we demonstrate its effectiveness compared to different baselines and previous methods. Additionally, we assess the generalization capabilities of our model on a new set of fashion items composed of upper-body clothes from different categories. To the best of our knowledge, we are the first to test virtual try-on architectures in this challenging experimental setting.

show abstract

Towards Photo-Realistic Virtual Try-On by Adaptively Generating↔Preserving Image Content

Cited by 235 publications

References 32 publications

DP-VTON: Toward Detail-Preserving Image-Based Virtual Try-on Network

DP-VTON: Toward Detail-Preserving Image-Based Virtual Try-on Network

TryOnGAN

VITON-GT: An Image-based Virtual Try-On Model with Geometric Transformations

Contact Info

Product

Resources

About