Identification of Violence Images Using Convolution Neural Network in Social Network

Abstract

In social networks some people may upload images that will contain violent content. In order to delete them,
we have to go back to the database and search the offending image among all and clear them. Instead of doing this we
use the software where we train it to learn features like bloodshed, by giving several datasets of violent images in order to
identify the images containing alcohol, violence, weapons etc. This method aims to filter violent images by using a
neural network to classify between the image categories. Leveraging the technique of image classification with neural
nets, this project aims to filter inappropriate image content. This can be done through Convolution Neural Network
(CNN) by processing an image.