SHOWMe

该数据集名为SHOWMe，包含了96个视频，每个视频都配有真实且详尽的手部与物体三维纹理网格注释，记录了与42种不同物体进行各种抓握类型及互动的场景。此外，该数据集专注于分割和刚性变换，以捕捉多样的手部与物体互动，从而实现精确的三维重建。规模上，它包含了来自15名参与者的96个视频序列，涉及42种不同的物体。该数据集的任务是进行三维手部与物体重建。

This dataset, named SHOWMe, consists of 96 videos, each paired with ground-truth and detailed 3D textured mesh annotations for hands and objects, documenting various grasp types and interaction scenarios involving 42 distinct objects. Furthermore, this dataset focuses on segmentation and rigid transformations to capture diverse hand-object interactions, thereby enabling accurate 3D reconstruction. In terms of scale, it includes 96 video sequences from 15 participants, involving 42 different objects. The core task of this dataset is 3D hand and object reconstruction.