Estimation Of 3d Body Shape And Clothing Measurements From Frontal-And Side-View Images

Abstract

The estimation of 3D human body shape and clothing measurements is crucial for virtual try-on and size recommendation problems in the fashion industry but has always been a challenging problem due to several conditions, such as lack of publicly available realistic datasets, ambiguity in multiple camera resolutions, and the undefinable human shape space. Existing works proposed various solutions to these problems but could not succeed in the industry adaptation because of complexity and restrictions. To solve the complexity and challenges, in this paper, we propose a simple yet effective architecture to estimate both shape and measures from frontal-and side-view images. We utilize silhouette segmentation from the two multi-view images and implement an auto-encoder network to learn low-dimensional features from segmented silhouettes. Then, we adopt a kernel-based regularized regression module to estimate the body shape and measurements. The experimental results show that the proposed method provides competitive results on the synthetic dataset, NOMO-3d-400-scans Dataset, and RGB Images of humans captured in different cameras.

The design, simulation, and adjustment of the Vietnam men’s Ao dai on the CLO3D software

“…The outcomes show that the size errors measured by 2D images were 10% less than those in manual measurements. Inferring body shapes through captured photographs is also the same experimental direction (3). The authors photographed the body's front and sides and subsequently applied the regression method to estimate the body shape and dimensions.…”