摘要

Visual Human Action Recognition is a universal hot topic of image processing, computer vision, pattern recognition, machine learning and artificial intelligence with wide applications in video surveillance, video retrieval, human-computer interaction, virtual reality, health care, etc. In this paper, we analyze the state-of-the-arts and advances of this field from perspectives of feature extraction, action recognition methods as well as benchmark datasets and competitions. In addition, the problems, difficulties, challenges and valuable future directions of human action recognition are presented.