Smart Safety Solutions: Utilizing MobileNet v2 for Violence Recognition to Enhance Women’s Safety

Violence against women is a harmful practice with very long-lasting consequences on the safety of women, which sometimes bring results in their physical and psychological wellbeing, affecting some abilities. Hitherto, all ages of women have been affected with some sort of crimes against their rights and are suffering in the worst way. The more we let things happen, the more they will grow. Even in a public space, women and girls face sexual violence in other dimensions; from unwelcome sexual remarks and gestures, to rape and femicide. It occurs on streets, in and around public transportation, schools, workplaces, public toilets, water and food distribution sites, and parks. Lastly, it bars women’s and girls’ free movement and their opportunity to attend school, work, and general life in public sphere. It has an adverse impact on their health and welfare, deprives them of access to services that are essential to them, and keeps them deprived of participating in cultural as well as entertaining activities.Through this research, We had focused on providing a safe public environment. This paper presents a safety solution aimed at enhancing women’s safety by utilizing deep learning techniques for real-time violence recognition. With the help of the effective MobileNet v2 architecture, a convolutional neural network optimized for embedded and mobile vision applications, the system can accurately and economically identify violent behavior in video data. The model is trained using violent and non-violent video data-set to ensure recognition in different environments. By integrating the suggested method into a mobile application, it provides a scalable and easily available tool for detecting violence in real time, alerting authorities, and ensuring prompt action.