Wearable Sensors and Real-Time System for Detecting violence using Artificial Intelligence

Aggressive activity in public spaces is a significant threat to personal safety and social cohesion. Cameras and other security devices have been mounted in various locations for public safety in recent years. Thousands of pieces of equipment are installed in public spaces, putting immense strain on security personnel. To classify incidents, almost all systems today need human review of the images, which is inefficient. Our proposed system is to develop an algorithm which detects violence in a given video frame. It first learns features and then trains on those learned features. It detects violence in given video and if violence is detected in frames, it will send an alert message. YOLOv5algorithmisfoundtobeabletoidentifyapersoninagi venvideo. Alongshort-termmemorynetwork (LSTM) is used to capture long-term dependency in the time domain.