Similarity measures of musical compositions based on content features
Автор: Pchelkin G.I.
Журнал: Форум молодых ученых @forum-nauka
Статья в выпуске: 5-2 (21), 2018 года.
Бесплатный доступ
The article is devoted to a review of methods and metrics for determining the similarity of audio compositions based on the characteristics of audio content. Features, such as low-level, rhythmic, tonal and high-level, are retrieved from the audio signal of the composition. The use of these characteristics as coordinates for determining the distance between compositions and for recommending similar musical compositions in the recommender system is reviewed.
Music, clustering, recommender system, content filtering, information retrieval, similarity measure, principal component analysis
Короткий адрес: https://sciup.org/140282653
IDR: 140282653
