UESTC-MMEA-CL

该数据集是首个用于持续自我中心活动识别的多模态数据集，它由参与者佩戴智能眼镜所收集的视频和传感器数据构建而成。此外，该数据集还建立了CEAR基准，并包含了三种基本的持续学习（CL）方法。数据规模覆盖了十位参与者的数据收集，其任务专注于持续的自我中心活动识别。

This dataset is the first multimodal dataset dedicated to continuous egocentric activity recognition. It is constructed from video and sensor data collected by participants wearing smart glasses. Additionally, this dataset establishes the CEAR benchmark and includes three fundamental Continual Learning (CL) methods. The dataset encompasses data from ten participants, with its tasks focused on continuous egocentric activity recognition.