Machine Recognition of Music Emotion

Abstract

The proliferation of MP3 players and the exploding amount of digital music content call for novel ways of music organization and retrieval to meet the ever-increasing demand for easy and effective information access. As almost every music piece is created to convey emotion, music organization and retrieval by emotion is a reasonable way of accessing music information. A good deal of effort has been made in the music information retrieval community to train a machine to automatically recognize the emotion of a music signal. A central issue of machine recognition of music emotion is the conceptualization of emotion and the associated emotion taxonomy. Different viewpoints on this issue have led to the proposal of different ways of emotion annotation, model training, and result visualization. This article provides a comprehensive review of the methods that have been proposed for music emotion recognition. Moreover, as music emotion recognition is still in its infancy, there are many open issues. We review the solutions that have been proposed to address these issues and conclude with suggestions for further research.

Continuous Music-Emotion Recognition Based on Electroencephalogram

“…The following EEG study using the same stimuli also confirmed the differences of cortical activities [9]. Considering the dynamic process of music emotion, a considerable amount of works in music emotion variation detection have been done in the last decade by utilizing musical features [10]. However, information from musical pieces does not always reflect the listener's emotional states.…”

Music Information Retrieval: Recent Developments and Applications

“…geographical location, influences) are linked to music context. Another example is semantic labels, which can be used to describe both the mood of a music piece and the emotion of a user as reviewed by Yang and Chen [305].…”

SMART: Semi-Supervised Music Emotion Recognition with Social Tagging

“…They build accurate models relying on a large amount of labeled data. In practice, however, labeled data is very hard to obtain in most genres except for Pop music [9] (automatic indexing systems can still be built given a genre [10]). This is because, human labor and "expert time" iare expensive.…”

“…To describe the audio features from a music emotion perpective, we consider several types of audio features, including timbre, pitch, rhythm, and loudness, which have been found to be effective features for describing music emotion [9].…”