Violence Content Detection in Videos

Social networks are not safe, because of the amount of violent content shared on these platforms. That’s why, filtering such content using a violence detection system powered by Artificial Intelligence is necessary and has become a growing research domain. In this paper, we propose a deep-learning-based approach to address this issue. We use a two layered model: First, a deep representation-based model that uses transfer learning concept to recognize violent content in a video. Second a text classifier to detect verbal violence using the audio cue. The result reports show that our approach is outperforming state-of-the art accuracies by learning most discriminating features achieving 90% as accuracy on the test set for physical violence detection and 89% for verbal violence detection.