IntelliMan_WP5_Grasping, Manipulation and Arm-Hand Coordination_T5.1_Data fusion and sensing technology_IEEEICECS2024_Data

The dataset was collected using a wearable sensing glove while grasping 28 daily-life objects. Five subjects participated in the experiment. For each trial, a subject grasped an object for 1 second and then released it. Each subject performed 10 trials per object, resulting in 50 trials per class and 1400 total recordings (28 objects × 50 trials). Each sample is a multichannel time series recorded at 200 Hz, with 200 time steps per trial (1 s × 200 samples/s) and 35 input channels in total, combining 5 FSR sensors and 5 IMUs (30 IMU channels). To evaluate the contribution of each sensing modality, the data were also split into two subsets: an IMU-only dataset (30 channels) and an FSR-only dataset (5 channels), enabling comparison of sensor configurations for object shape recognition.本数据集采用可穿戴传感手套（wearable sensing glove）采集，采集场景为抓取28种日常物品。本次实验共有5名受试者参与。每次试验中，受试者抓取物品并保持1秒后松开。每名受试者针对每种物品完成10次试验，因此每类别含50次试验，总录制样本量为1400条（28个物品 × 50次试验）。 每条样本为采样率200赫兹的多通道时间序列，每次试验包含200个时间步（即1秒 × 200个采样/秒），总共有35个输入通道，其中包含5个力敏电阻（Force Sensitive Resistor, FSR）传感器与5个惯性测量单元（Inertial Measurement Unit, IMU），后者共计提供30个IMU通道。为评估每种传感模态的贡献，数据集还被划分为两个子集：仅IMU数据集（30个通道）与仅FSR数据集（5个通道），可用于对比不同传感器配置在物品形状识别任务中的表现。