monster-monash/Skoda
收藏资源简介:
Skoda数据集是MONSTER项目的一部分,专注于车辆维护场景中10种特定的操作手势。该数据集旨在研究手势的故障容错性、传感器数量对性能的可扩展性以及功率性能管理。数据集由10个操作手势类别组成,使用2 x 10个USB 3D加速度传感器在左右上下臂上记录。传感器具有大约98 Hz的高采样率,确保了动作捕捉的精确性。数据集包含大约3小时的录制时间,共记录了每个手势70次,总计14,117个时间序列,每个序列长度为100,代表大约每秒98 Hz的数据。
The Skoda dataset is a part of the MONSTER project, focusing on 10 specific manipulative gestures in a car maintenance scenario. The dataset aims to investigate various aspects of gestures, such as fault resilience, performance scalability with the number of sensors, and power performance management. It consists of 10 classes of manipulative gestures recorded using 2 x 10 USB 3D acceleration sensors placed on the left and right upper and lower arms. The sensors operate at a high sampling rate of approximately 98 Hz, ensuring precise movement capture. The dataset includes approximately 3 hours of recording time, with each gesture recorded 70 times, totaling 14,117 time series each of length 100, representing about one second of data per time series at 98 Hz.




