Wearable SELD dataset

Wearable SELD数据集是由早稻田大学和日本电信电话公司合作创建的，旨在通过穿戴设备实现声音事件定位与检测。该数据集包含24个麦克风记录的数据，这些麦克风被安装在头部和躯干模拟器上，模拟眼镜、耳机等穿戴设备。数据集创建过程中，使用了多种录音条件和环境，以模拟真实世界的声音场景。该数据集主要用于开发和评估穿戴式声音事件定位与检测系统，特别是在行人行走等动态环境中的应用。

Wearable SELD dataset was jointly created by Waseda University and Nippon Telegraph and Telephone (NTT) Corporation, aiming to achieve sound event localization and detection (SELD) via wearable devices. This dataset contains recordings from 24 microphones mounted on a head-and-torso simulator, which simulates wearable devices such as glasses and headphones. During the dataset development, various recording conditions and environments were adopted to simulate real-world acoustic scenes. This dataset is primarily used for developing and evaluating wearable sound event localization and detection systems, especially for applications in dynamic scenarios like pedestrian walking.