Application of Deep Learning for Image Sequence Classification

Abstract

Understanding human behavior from videos is a difficult challenge in computer vision. Computers are already pretty competent at recognizing objects in static photos, and such technology is frequently utilized today, but doing so in moving images is far more difficult. Identifying human behavior in films is a key step in creating smarter, gentler robots. Identifying the human activity in motion neural network plays a crucial role. In this study, we proposed a convolutional neural network to identify human activities in video. This algorithm helps in many sectors like anti-terrorists, anti-crime securities, and as well as surveillance, life logging and assistance.