EHUNAM, a WiFi CSI-based dataset for human and machine sensing

This is a comprehensive WiFi Channel State Information (CSI) dataset collected using different device configurations for multiple WiFi sensing applications such as People Counting (PC), Human Activity Recognition (HAR), Machine Recognition (MR), and Machine Activity Recognition (MAR). To ensure high accuracy in new locations, data was collected over 23 days in eight different environments, including an industrial scenario with 21 people and nine machines that can also perform activities simultaneously. The dataset also includes the Summary.xlsx and Summary.mat files, which provide details about each of the dataset files.

本数据集为一套完备的WiFi信道状态信息（WiFi Channel State Information，CSI）数据集，针对人员计数（People Counting，PC）、人体活动识别（Human Activity Recognition，HAR）、设备识别（Machine Recognition，MR）与设备活动识别（Machine Activity Recognition，MAR）等多种WiFi感知应用场景，采用不同设备配置采集所得。为保障新部署场景下的检测精度，研究团队在8种不同环境中历时23天完成数据采集，其中包含一个可支持21名人员与9台设备同时开展活动的工业场景。本数据集还附带Summary.xlsx与Summary.mat文件，可用于详细说明各数据集文件的相关信息。