Deep Learning-Based Smart Surveillance System

The growing usage of security cameras in smart cities to enable round-the-clock surveillance has allowed researchers to analyze a vast amount of data. A better security system is required in other monitoring industries to stop any casualties that can result in monetary, societal, and ecological loss. Automatic violence detection is crucial for prompt response and can effectively help the relevant authorities. Intelligent surveillance technologies are desperately needed to keep an eye on people and identify their aggressive behavior in public places like banks, hospitals, shopping malls, and train stations, among others. As a result, computer vision researchers are increasingly interested in the identification of violent behavior. For the sake of public safety, it is imperative that methods for spotting violence in films be successful and effective. For the identification of these actions, several handmade and deep learning feature-based algorithms have been developed during the last several years. This paper describes the emergency warning system we designed using deep learning. The suggested technique is solely appropriate for violence detection.