BUFF(Bodies Under Flowing Fashion)

可靠的3D身体形状评估对于许多应用 (例如健身监控，虚拟试衣等) 至关重要。但是在极小的衣服下扫描身体是这项技术研究的一个主要问题。有关研究人员提出通过使用三维扫描方法来解决这一技术问题，并为此提出了BUFF数据集进行定量评估。 BUFF数据集的全称是Bodies Under流动时尚，是一个高质量的四维数据集。共有3名男性和2名女性参加了数据收集。他们穿着两种款式的服装: t恤和长裤或足球服，并做出了三种不同的动作: 臀部，tilt_twist_left，shouders_mill。Reliable 3D body shape assessment is critical for numerous applications such as fitness monitoring and virtual try-on. However, scanning the human body while wearing extremely tight clothing poses a major challenge in this technical research area. Relevant researchers have proposed to address this technical issue using 3D scanning methods, and developed the BUFF dataset for quantitative evaluation. The full name of the BUFF dataset is Bodies Under Flow Fashion, which is a high-quality four-dimensional dataset. A total of 3 male and 2 female participants took part in the data collection. They wore two styles of clothing: T-shirts and trousers, or football jerseys, and performed three distinct actions: hip movement, tilt_twist_left, and shouders_mill.

搜集汇总

背景与挑战

背景概述

BUFF数据集是一个高质量的四维数据集，旨在通过三维扫描方法解决在极小衣服下评估3D身体形状的技术问题。该数据集包含3名男性和2名女性参与者，他们穿着t恤和长裤或足球服，并执行了三种不同的动作，由纽约大学于2017年发布。