putEMG
收藏资源简介:
putEMG数据集是由波兹南理工大学的控制、机器人与信息工程研究所创建,专注于表面肌电图手势识别。该数据集包含44名健康受试者的数据,涵盖8种手势(包括3种全手势、4种捏合动作和空闲状态)。数据集通过24通道表面肌电图记录,结合RGB视频流和深度相机图像进行手部运动跟踪。此外,数据集还提供了示例处理脚本,便于快速采用。该数据集适用于开发和测试用户不变的分类方法,以及评估电极定位概念和分类方法的鲁棒性。
The putEMG dataset was developed by the Institute of Control, Robotics and Information Engineering at Poznań University of Technology, with a focus on surface electromyography (sEMG)-based gesture recognition. It comprises data collected from 44 healthy subjects, covering 8 gesture categories including 3 full-hand gestures, 4 pinch movements, and a resting state. The dataset is recorded via a 24-channel surface EMG system, and is supplemented with RGB video streams and depth camera images to enable hand motion tracking. Additionally, sample processing scripts are provided to facilitate rapid adoption by researchers. This dataset is suitable for developing and testing user-invariant classification methods, as well as evaluating the robustness of electrode positioning concepts and classification algorithms.




