Emotion based mood enhancing music recommendation

摘要

Music is one of the most effective media as it can instill deep feelings and swamp listeners with subliminal messages. It deftly plays with our emotions which in turn affect our mood. Books, movies and television dramas are a few other media but, in contrast to these, music delivers its message in mere moments. It can aid us when we are feeling low and empower us. When we listen to sad songs, we tend to feel a decline in mood. When we listen to happy songs, we feel happier. Manual classification of songs based on mood, for making of a playlist, is time consuming and labour intensive. Our paper proposes a system `EmoPlayer', an Android application, which help to minimize these efforts by suggesting the user a list of songs based on his current emotion. The system captures user's image using camera and detects his face. It then detects the emotion and makes a list of songs which will enhance his mood as the songs keep playing. EmoPlayer uses Viola Jones algorithm for face detection and Fisherfaces classifier for emotion classification.