Applications such as video surveillance, robotics, source selection, and video indexing often require the recognition of actions based on the motion of different actors in a video. Certain applications may require assigning activities to several predefined classes, while others may rely on the detection of abnormal or infrequent activities. In this summary we provide a survey of dominant models and methods and discuss recent developments in this domain. We briefly describe two recent contributions: joint level feature and sequence learning, as well as space-time graph matching.