IDENTIFYING WOMEN HARASSMENT IN PUBLIC CCTV CAMERASUSINGARTIFICIALINTELLIGENCE TECHNIQUES
Keywords:
MIL – Multiple Instance Learning, SVM – Support Vector Machine, C3D – Cafe to support3-DimensionalConvolutionNetworks,HOG–HistogramofOrientedGradient,AI – ArtificialIntelligence.Abstract
The society probably living in the worst time. Our current society has ever seen in terms ofwomen security. The projects aim is to make women experience as strong as ever and sufficiently ableto fight the parasites of our general public. Our project is an idea that makes every place safer forwomen.Nowadaysthenewsaboutwomenharassmentismorethantheiraccomplishments.Thepurposeof the proposed system is to decrease the incidence of sexual harassment faced by women in their ifeexperiencesandtoincreasewomen'sself-confidencetostepoutintotheirofficeduringnightsanddesiretouse publicspacesinthecity especially duringnighttime. Womenfeelunsafe at bizarretimes to be moving alone. There are many applications built for the safety of women.Even there aremany uneducated women those who do not have android phones or do not know how to handle it. Theproposed system aims to provide safety for women who are in trouble. Surely there will be CCTVcameras in every public place. This model does not require any special hardware and it uses algorithmfor the precaution of women. The algorithm is then applied on the CCTV cameras and then the genderof the person will be predicted. The alert message will be sent to the nearby police station. This projectis designedfor the conveniences of allwomen.
