MSRC-12 (MSRC-12 Kinect Gesture Dataset)

Microsoft Research Cambridge-12 Kinect 手势数据集由人体运动序列组成，表示为身体部位位置，以及系统识别的相关手势。该数据集包括 594 个序列和 719,359 帧 - 大约 6 小时 40 分钟 - 从 30 个人执行 12 个手势收集。总共有 6,244 个手势实例。运动文件包含使用 Kinect 姿势估计管道估计的 20 个关节的轨迹。身体姿势以 30Hz 的采样率捕获，关节位置的精度约为 2 厘米。

The Microsoft Research Cambridge-12 Kinect Gesture Dataset comprises human motion sequences represented by body joint positions and associated gestures identified by the corresponding system. The dataset includes 594 sequences and 719,359 frames, amounting to approximately 6 hours and 40 minutes of data collected from 30 individuals performing 12 distinct gestures. A total of 6,244 gesture instances are contained within the dataset. The motion files hold the trajectories of 20 joints estimated using the Kinect pose estimation pipeline. Body poses are captured at a sampling rate of 30 Hz, with the joint positions having an accuracy of approximately 2 centimeters.

搜集汇总

背景与挑战

背景概述

MSRC-12是一个基于Kinect的手势数据集，包含30个人执行的12种手势的594个序列，总计6,244个手势实例。数据以30Hz采样率捕获，包含20个关节的轨迹信息，精度约为2厘米，适用于动作检测和2D关键点检测等研究。