Music has been a popular way to convey and comprehend human emotions, and it is a powerful medium of expression. To enhance emotional well-being, this paper involves identifying an individual's emotional state and playing music that matches their current feelings. Existing models of emotion-based music recommendation are based on various algorithms like Principal Component Analysis (PCA), Support Vector Machines (SVM), Artificial Neural Network (ANN), etc. However, those models have not yielded optimal results and are still being pursued. This paper aims to improve these limitations by proposing a simple, user-friendly Machine Learning (ML) system that can determine human emotional state. This research endeavour is centered around training an ML model with extensive datasets of facial emotions as input training data from the user. This system uses physiological data from the user's face to identify emotional patterns and affinities and suggests playlists, music tracks, and albums that match the recognized emotion. These models intend to capture the subtleties of various emotional states, such as joy, sadness, neutrality, rock, and surprise. This paper is a significant advance in the use of machine learning and music, presenting an innovative method for deeply linking music to human emotions.