Emotion Based Music Recommendation System Using OpenCV and Machine Learning Technique

Emotion plays a major role in influencing human music preferences. Traditional music recommendation systems mainly rely on listening history, ratings, and playlists, but they often fail to understand the user’s real-time emotional state. This paper presents an Emotion Based Music Recommendation System using OpenCV, MediaPipe, Machine Learning, and Streamlit that detects user emotions through facial expressions and recommends suitable music accordingly. The system captures facial images through a webcam, extracts facial and hand landmarks using MediaPipe Holistic, and predicts emotions using a trained Convolutional Neural Network model. Based on the predicted emotion and user preferences such as language and singer, the system automatically recommends songs using YouTube automation and local music playback. The proposed system demonstrates the practical application of effective computing and intelligent recommendation systems in entertainment platforms.