Virtual Try-on Network With Attribute Transformation and Local Rendering

Xu, Jun; Pu, Yuanyuan; Nie, Rencan; Xu, Dan; Zhao, Zhengpeng; Qian, Weiping

doi:10.1109/tmm.2021.3070972

Cited by 16 publications

(5 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the tenth and eleventh rows, in the complex posture of arm-occlusions, our method successfully puts the clothes in the corresponding position and preserve the arm details, while the competition methods either cover the arms with clothes or generate fake arms. on methods (VITON [13], CP-VTON+ [24], GarmentGAN [29], SieveNet [17], VTNFP [42], VTON-ATLR [40]) without public code. Note that the results of the competition methods are obtained by their papers.…”

Section: Qualitative Resultsmentioning

confidence: 99%

“…The authors in [6,17,29,42] proposed additional semantic prediction module to improve the overall try-on performance . The authors in [4,8,40,41] presented various methods to preserve the details of non-target region. However, the limitation of the above methods is that the human parser is still needed during testing.…”

Section: Virtual Try-onmentioning

confidence: 99%

“…However, the available datasets [13] is an unpaired data (clothes and a person wearing the clothes), and it cannot access to the ground-truth after clothing transfer. To solve this issue, the existing methods [4,6,8,13,16,17,24,29,38,[40][41][42] cast virtual try-on into a reconstruction task. They mask the clothing region of a person image with a pre-trained human parser, and then reconstruct the clothing region of the masked person image by using the same clothes for supervised learning.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

UF-VTON: Toward User-Friendly Virtual Try-On Network

Chang

Peng

et al. 2022

Proceedings of the 2022 International Conference on Multimedia Retrieval

View full text Add to dashboard Cite

Image-based virtual try-on aims to transfer a clothes onto a person while preserving both person's and cloth's attributes. However, the existing methods to realize this task require a target clothes, which cannot be obtained in most cases. To address this issue, we propose a novel user-friendly virtual try-on network (UF-VTON), which only requires a person image and an image of another person wearing a target clothes to generate a result of the person wearing the target clothes. Specifically, we adopt a knowledge distillation scheme to construct a new triple dataset for supervised learning, propose a new three-step pipeline (coarse synthesis, clothing alignment, CCS CONCEPTS• Computing methodologies → Image-based rendering; Texturing.

show abstract

Section: Qualitative Resultsmentioning

confidence: 99%

Section: Virtual Try-onmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

UF-VTON: Toward User-Friendly Virtual Try-On Network

Chang

Peng

et al. 2022

Proceedings of the 2022 International Conference on Multimedia Retrieval

View full text Add to dashboard Cite

show abstract

“…Virtual reality technology is becoming more important in the apparel industry for various reasons [12][13]. Virtual reality and augmented reality are now more accessible than before, prices have dropped so that many fashion houses can afford them, and interesting applications can be created to help brands make the most of various fashion ideas [14]. Yu, J. et al The study of clothing ethnic identity and totemic area extraction was explored through visual interaction, and the importance of virtual reality technology for clothing design was emphasized [15].…”

Section: Introductionmentioning

confidence: 99%

Research on the design and presentation of traditional ethnic costumes based on virtual reality technology

2023

Applied Mathematics and Nonlinear Sciences

View full text Add to dashboard Cite

In this paper, the 3D model of the human body is drawn, and the 3D transformation in the 3D garment piece design is realized by using the inverse mapping of 2D screen coordinates to 3D model coordinates. Then, the technique of virtual presentation of traditional ethnic clothing is studied, alternating the left and in-eye views and using the overlap of calculation to optimize the calculation to achieve a three-dimensional display. The virtual design and presentation effects of traditional ethnic costumes were assessed through a comprehensive comparison, sensory experience, interactive experience, emotional experience, and cognitive experience. The results showed that the regression coefficients of cognitive experience, sensory experience, interactive experience, and emotional experience of the virtual design and presentation of costumes were 0.98, 0.875, 0.993, and 0.996, respectively, which were 0.506, 0.36, 0.11, 0.161, and 0.048 more than the regression coefficients of the traditional display method in each dimension. This study provides the theoretical basis for combining traditional costumes with virtual technology.

show abstract

“…Therefore, this method avoids a series of problems caused by traditional warping methods, such as mismatch between body parts and clothing, difficulty in learning complex postures, and body parts covering clothing. • We break the limitations of paired and unpaired dataset and achieve state-of-the-art performance on the highresolution [20] datasets. For the SCM and TGM, their garment input can be either the in-store tiled garment image or the model image, which does not affect the final result.…”

Section: Introductionmentioning

confidence: 99%

A Novel MS-MeMBer Filter for Extended Targets Tracking

Zhang

Sun

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Conventional multi-sensor multi-target multi-Bernoulli (MS-MeMBer) filters are based on the assumption that each target produces at most one measurement per time step. However, this assumption is not always reasonable in practice as an extended target can generate multiple measurements per step due to the recent improvement in the sensor resolution. In this case, a potential estimation bias may occur in the current MS-MeMBer filters. Therefore, a novel extended target MS-MeMBer filter and its Gaussian inverse Wishart mixture implementation are given in this paper. Specifically, we modify the update process of the MS-MeMBer filter by assuming that the generation of extended target measurements follows an approximate Poisson-Body model. Simulation results validate that the proposed filter can effectively estimate the shape and position of the extended target. INDEX TERMS Multi-sensor multi-target multi-Bernoulli filter, extended target, approximate Poissonbody, Gaussian inverse Wishart.

show abstract

Virtual Try-on Network With Attribute Transformation and Local Rendering

Cited by 16 publications

References 36 publications

UF-VTON: Toward User-Friendly Virtual Try-On Network

UF-VTON: Toward User-Friendly Virtual Try-On Network

Research on the design and presentation of traditional ethnic costumes based on virtual reality technology

A Novel MS-MeMBer Filter for Extended Targets Tracking

Contact Info

Product

Resources

About