A set of advanced approaches and models on human action, activity, gesture, behavior and related aspects are summarized in this note. There are a number of challenges on these domains, and some of these are addressed here. Notably, Video-based human activity recognition, sensor-based activity analysis, skeleton-based activity recognition, assisted daily living, anomaly detection, facial expression and emotion analysis, gesture and sign language are highlighted in the works. This special issue also introduced six new datasets, while exploring a total of 55 different datasets in 21 papers. However, there are still various challenges that we need to solve in the future. Multi-modality, action quality assessment, real-time applications, and activity and behavior computing on edge devices are some of the dominant future challenges to deal with. We firmly believe that the original works and ideas presented in this special issue will serve as helpful references for the relevant research communities in the journey towards a brighter future.

本说明中总结了一组关于人类行为、活动、手势、行为和相关方面的高级方法和模型。这些领域存在许多挑战，其中一些在此处得到解决。值得注意的是，基于视频的人体活动识别、基于传感器的活动分析、基于骨骼的活动识别、辅助日常生活、异常检测、面部表情和情绪分析、手势和手语在作品中得到了突出体现。本期特刊还介绍了六个新数据集，同时在 21 篇论文中探索了总共 55 个不同的数据集。但是，我们未来仍然需要解决各种挑战。边缘设备上的多模态、动作质量评估、实时应用程序以及活动和行为计算是未来需要应对的一些主要挑战。