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

Chang, Yuan; Peng, Tao; He, Ruhan; Hu, Xinrong; Liu, Junping; Zhang, Zili; Jiang, Minghua

doi:10.1145/3512527.3531387

Cited by 4 publications

(1 citation statement)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In response to this challenge, virtual try-on technology optimizes simulations that resemble real-world try-on scenarios. Existing research [1][2][3][4][5][6]predominantly focuses on virtual try-on solutions utilizing two-dimensional images to provide feedback on the chosen garment's appearance. Nevertheless, persistent challenges exist in accurately adjusting and aligning garments when the character's pose varies significantly and is affected by occlusion.…”

Section: Introductionmentioning

confidence: 99%

GIC-Flow: Appearance Flow Estimation Via Global Information Correlation for Virtual Try-On under large deformation

Zhang,

Zhan,

Sun

et al. 2024

Preprint

View full text Add to dashboard Cite

The primary aim of image-based virtual try-on is to seamlessly deform the target garment image to align with the human body. Due to the inherent non-rigid nature of garments, current methods prioritize flexible deformation through appearance flow with high degrees of freedom. However, existing appearance flow estimation methods focus solely on the correlation of local feature information. While this strategy successfully avoids the extensive computational effort associated with directly computing the global information correlation of feature maps, it leads to challenges in garments adapting to large deformation scenarios. To address these limitations, we propose the GIC-Flow framework, which obtains appearance flow by calculating the global information correlation while reducing computational regression. Specifically, our proposed Global Streak Information Matching Module is designed to decompose the appearance flow into horizontal and vertical vectors, effectively propagating global information in both directions. This innovative approach significantly diminishes 1 computational requirements, contributing to an enhanced and efficient process. Additionally, to ensure the accurate deformation of local texture in garments, we propose the Global Aggregate Information Matching Module to aggregate information from nearest neighbors before computing the global correlation, and to enhance weak semantic information. Comprehensive experiments conducted with our method on the VITON and VITON-HD datasets demonstrate GIC-Flow outperforms existing state-of-the-art algorithms, particularly in cases involving complex clothing deformation.

show abstract