Deep Learning-Based Video Action Recognition: A Review

Video action recognition refers to the process of recognizing human actions in a video. The field of action recognition encompasses a wide variety of high-impact societal applications, including video surveillance, gaming, human-computer interaction, web-video search and retrieval, multimedia understanding, medical diagnosis and many other fields. In this research study, we comprehensively review deep learning-based video action recognition methods with special emphasis on methods based on convolutional neural networks. We also address future research directions and limitations of deep learning-based video action recognition methods. This review paper serves as a guide for future research in this domain.