A Comparative Study and Review for Human Action Recognition

Abstract: This paper presents a study and review of state-of-the-art research papers on human action recognition. One of the challenging issue is the process of recognizing and understanding of human actions from videos owing to large variations in human appearance, pose changes, scale changes etc. The most important approach for human action recognition is to extract features from videos as representations.It is a main area of computer vision approach.The main applications include surveillance systems, patient monitoring systems, and a number of systems that involve interactions between persons and electronic devices such as human-computer interfaces. Almost all applications require an automatic recognition of high level activities. A brief overview of various state-of-the-art research papers on human action recognition is discussed here.Action Recognition based on Sparse Representation is one of the latest method for a higher recognition accuracy and improving the performance.For this any Datasets can be used Weizmann Human Action Dataset, UCF Sports dataset, Ballet Dataset.

Keywords: Background subtraction, segmentation, K-SVD Random Projection, Gaussian Smoothing.