Calculating Screen Time of Characters in a Video Using Convolutional Neural Networks

Abstract

In this growing world of digital technology, images and videos are one of the richest sources of data on the internet. Several technologies have emerged which help the users to deal with the image and video data. In this paper, we propose a methodology to calculate the screen time of characters appearing in the video by using the classification-count method.Supervised Learning Methodology with Convolutional Neural Network as the classifier is used in this paper. The proposed methodology is implemented on a series of Tom and Jerry videos available publicly on the internet. Experimental results show that the proposed methodology has shown good accuracy.