Crowd Video Event Classification using Convolutional Neural Network

Abstract

Crowd Event Classification in videos is an important and challenging task in computer vision based systems. The Crowd Event Classification system recognizes a large number of video events. The decisive of the model is a difficult task in the event classification. The event classification model has generalization capability on works with a higher number of videos. The embodiment of Deep Learning in video event classification derives powerful and distinguishes feature portrayals. The features of events are extracted from raw data through massive videos with effective and efficient detection. The Convolutional Neural Network (CNN) has been established as a powerful classification model for event recognition problems. A higher quality of new dataset of 3000 frames collected from youtube videos belonging to four classes of crowd events namely Marriage, Cricket, Jallikkattu, and Shopping mall The system has used two Deep CNN infrastructures are namely baseline and VGG16, which detects predefined events and provides temporal evidence. The CNN Model automatically tests input video frames and detects the events of centrality at the video. The CNN extracts the video events features from the video input frames and distinguishes the events name correctly. The system shows more improved 100% results compare with each other models.