An Xception Model Based Real-time Violence Detection
Narenthira Kumar Appavu, C. Nelson Kennedy Babu · 2023
Nowadays the usage of CCTV surveillance is increasing. The number of cases of violence in public places like airports, railway stations, bus stops, shopping malls, restaurant, sports arenas and party halls have only been discovered after the fact. It is too late. Our objective be located to develop a complete structure capable of video analysis in real-time manner, which can identify the existence of violence from input video such as cctv and report the violence to concerned authorities such as higher authorities, police to take necessary action. In the proposed work using CNN for feature extraction and LSTM Interpreting features in chronological order. The distinct Xception system manner, we Achieving an effective solution for analysis video in real-time fashion through violent scenes obtained from the video allows us to legally monitor the situation with the relevant authorities. A mobile app, Telegram bot that can report the violent event Immediately to concern authorities.