Human Action Recognition In Streaming Video Using Deep Learning
Human action recognition in streaming video is an open issue in surveillance monitoring system in the field of Artificial Intelligence. To protect and prevent from robbery, chain snatching issue surveillance monitoring system is necessary to the society. In this paper we discussed about a novel architecture of Deep learning based CNN is used to address the above issue. Human Action frame prediction model is used to predict the various action of human in streaming video. UCF101 data set is used to predict the various actions with learning rate of 0.01. The proposed approach yields 72.34% accuracy of identifying the human actions compared to various traditional models.