A Novel Framework for Women's Safety Using Computer Vision and Machine Learning
摘要
You can find that women are equally participating across different sectors like business, health, education, etc. irrespective of regions in the world. Sometimes, women have to face harassment or embarrassing situations in public places, offices, etc. The advancement of computer vision and machine learning can be used to automatically detect and protect women from such situations. In this paper, a novel framework is proposed for women’s safety using computer vision and machine learning. The proposed framework will be able to automatically detect if the woman is in trouble like harassment. When the AI module detects such situations. It will start playing an alarm message in that locality. The framework will also send a help request to the nearby police station along with the location information and the footage of the situation. The success of this framework mainly depends on the AI module, i.e. accurate classification of gender and expression of the persons present in the situation. The overall accuracy for gender and expression classification is 95.37 and 83.20%, respectively.