Moody.ai: An Adaptive Activity Recommendation System Based on Emotion Detection

According to a report published by the World Health Organization in 2021, depression affects an estimated 5% of individuals worldwide. Most people who experience sadness or anxiety try to ignore the problem and get on with their lives. Apart from this, people experiencing specific moods frequently don’t know what to do in order to boost or reduce the feelings associated with those moods. A person’s good mood is ruined by the uncertainty and constraints surrounding what to do, which intensifies feelings of perplexity, anger, melancholy, discomfort, etc. Until they look in the mirror while experiencing an episode brought on by the bad mood, most people find it difficult to accept when they’re in a terrible mood. So, based on an Emotion Detection Algorithm, this paper proposes an adaptive Activity Recommendation System. A publicly available dataset of facial images called FER-2013 is used in the proposed emotion recognition model, which is based on the Visual Geometry Group (VGG-16) Convolutional Neural Network (CNN) model. The suggested recommendation method takes into account entertainment-related activities such as watching movies, reading books, playing sports, listening to music, and other enjoyable pursuits. The recommendation engine utilises a datasets of films, songs, books, sports, and other entertainment activities. The accuracy of the suggested emotion recognition model is 90%.