Description of Invention
Nowadays, many blind individuals are facing difficulties in recognising faces and facial expressions, which can have significant impact on their social interactions and ability to navigate the world around them. This project aims to address the difficulties faced by blind individuals in recognising faces and facial expressions and to improve the social and emotional well-being of blind individuals and enhance their ability to navigate the world around them. The system uses Dlib with HOG and SVM for face detection, ResNet50 model for face recognition with 99.38% for the accuracy and CNN for facial expressions recognition with 70% for training accuracy and around 60% for validation accuracy.