Synthetic Human Model Dataset
收藏资源简介:
Synthetic Human Model Dataset是由澳大利亚联邦科学与工业研究组织数据61创建的合成数据集,专门用于评估基于RGB-D相机的非刚性3D人体重建技术。该数据集包含七个不同的运动序列,每个序列提供每帧的地面真实几何和骨骼信息,以及皮肤权重数据。数据集的创建过程涉及精确的骨骼关节捕捉,以支持高帧间非刚性运动的重建。该数据集主要应用于计算机视觉和机器人领域,旨在解决动态场景下的3D重建问题,特别是在人体运动跟踪和重建方面。
The Synthetic Human Model Dataset is a synthetic dataset created by Data61 at the Commonwealth Scientific and Industrial Research Organisation (CSIRO) of Australia, specifically designed for evaluating non-rigid 3D human reconstruction technologies based on RGB-D cameras. This dataset contains seven distinct motion sequences, where each sequence provides per-frame ground-truth geometry, skeletal information, and skin weight data. The development of this dataset involves precise skeletal joint capture to support the reconstruction of high-fidelity inter-frame non-rigid human motions. It is primarily applied in the fields of computer vision and robotics, aiming to address 3D reconstruction challenges in dynamic scenarios, particularly for human motion tracking and reconstruction.




