In previous chapters, readers have learned a variety of analysis techniques for images, including object detection, instance segmentation, and even human pose estimation. However, in the real world, everything is in motion, so it is necessary to master techniques for video streams. In this chapter, we will try to understand this world of motion, specifically, we will introduce human action recognition. As the name suggests, human action recognition refers to the recognition of human actions in videos, which can infer the intentions and states of the characters, providing important information for understanding video content.