monster-monash/USCActivity

USCActivity数据集是由运动节点设备收集的数据组成的，该设备包括从佩戴在身上的3轴加速度计和陀螺仪传感器获得的六个读数。这个数据集包含了14名男女受试者（各有7名）的样本，这些受试者具有特定的身体特征和年龄。传感器数据以100Hz的频率采样，数据集中的每个时间步都被标记为12个活动类别之一。该处理后的数据集由56,228个长度为100的多变量时间序列组成，每个时间序列代表100Hz采样率下的一秒钟数据。由于传感器的放置和活动类别的多样性，该数据集在学习特征表示和分割方面具有挑战性。

The USCActivity dataset consists of data collected from a Motion-Node device, which includes six readings from a body-worn 3-axis accelerometer and gyroscope sensor. The dataset contains samples from 14 male and female subjects with equal distribution (7 each) and specific physical characteristics and ages. The sensor data is sampled at a rate of 100 Hz, and each time-step in the dataset is labeled with one of 12 activity classes. The processed dataset consists of 56,228 multivariate time series each of length 100, representing one second of data at 100 Hz. The dataset presents a challenge in learning feature representation and segmentation due to the placement of the sensors and the variability in activity classes.