Unite the People

团结人是用于3D人体姿势估计的图像数据集。这些图像来自利兹体育姿势数据集及其扩展，以及MPII人类姿势数据集。每个图像都有不同类型的注释，例如分割标签，pose或3D，图像中的人体模型分为31个部分和91个关键点。

Tuanjie Human is an image dataset designed for 3D human pose estimation. Its source images are collected from the Leeds Sports Pose Dataset and its extended version, as well as the MPII Human Pose Dataset. Each image is equipped with multiple types of annotations, such as segmentation labels, pose annotations, or 3D pose annotations. The human body models in these images are divided into 31 body segments and 91 key points.

搜集汇总

背景与挑战

背景概述

Unite the People是一个用于3D人体姿势估计的图像数据集，其图像来源于利兹体育姿势数据集及其扩展以及MPII人类姿势数据集。该数据集提供了分割标签和姿势（含3D）等注释，人体模型被划分为31个部分和91个关键点，由Bernstein Center for Computational Neuroscience, Tubingen于2017年发布。