Audio Features for Music Emotion Recognition: A Survey

Abstract

The design of meaningful audio features is a key need to advance the state-of-the-art in Music Emotion Recognition (MER). This work presents a survey on the existing emotionally-relevant computational audio features, supported by the music psychology literature on the relations between eight musical dimensions (melody, harmony, rhythm, dynamics, tone color, expressivity, texture and form) and specific emotions. Based on this review, current gaps and needs are identified and strategies for future research on feature engineering for MER are proposed, namely ideas for computational audio features that capture elements of musical form, texture and expressivity that should be further researched. Finally, although the focus of this article is on classical feature engineering methodologies (based on handcrafted features), perspectives on deep learning-based approaches are discussed.

An Analysis of the Possible Significance of Music Education on the Practice of a Performer

“…Still, he understands why in retrospect because he had not been trained to perform as effectively as he does now. In this statement, he shows both the appreciation for the lack of musical education and its impact on the performer and the contrast that having a good grounding in musical education has on a performer [17]. Now, bearing in mind the career that John Legend has gone on to have after undergoing further training, establishing himself as one of the most influential singers of the past decade, the importance of musical education cannot be overstated.…”

“…The structure and perception of music give an insight into the learning and information retention processes because of human creative cognition, which is communicated as a component of an oral tradition [17]. This can be referenced in such scenarios as the best length of a melody for optimal learning and recall, which devices could help in remembrance, and the roles of rhythm and hierarchical organization in memory [17].…”

“…Music and emotion are related closely in many ways [17]. However, music-related emotions are difficult to pinpoint and describe whether they are common emotions experienced in everyday life.…”

“…The text version of a music composition/component is a set of sounds aimed at arranging the modal-rhythmic intonations aesthetically. This representation features sensual signals, cultural meanings, and the actualization of personal semantic experiences that create art emotions [17]. The mind looks for ways to complete the intrinsic world of work in any musical activity.…”

Music emotion recognition using recurrent neural networks and pretrained models

“…A comprehensive review of the current emotionally-relevant computational audio features used in MER was presented by Panda et al (2020). They show the relations between eight musical dimensions (melody, harmony, rhythm, dynamics, timbre, expressivity, texture, and form) and specific emotions.…”