[논문]Music Similarity Search Based on Music Emotion Classification

Kim, Hyoung-Gook; Kim, Jang-Heon

[국내논문] Music Similarity Search Based on Music Emotion Classification 원문보기

This paper presents an efficient algorithm to retrieve similar music files from a large archive of digital music database. Users are able to navigate and discover new music files which sound similar to a given query music file by searching for the archive. Since most of the methods for finding similar music files from a large database requires on computing the distance between a given query music file and every music file in the database, they are very time-consuming procedures. By measuring the acoustic distance between the pre-classified music files with the same type of emotion, the proposed method significantly speeds up the search process and increases the precision in comparison with the brute-force method.

* AI 자동 식별 결과로 적합하지 않은 문장이 있을 수 있으니, 이용에 유의하시기 바랍니다.

문제 정의

In this paper, we propose an efficient music similarity 응earch system based on the results of music emotion classification. A new tempo feature based on MDCT and wavelet transformation is combined with conventional timbre features in order to significantly improve the precision of music classification and similarity results.

가설 설정

1) If the query music file is classified to sad emotion class, the distance between the query signature and every signature with sad emotion subset in the music database is computed.2) The confusion set of the sad emotion subsets is the calm emotion subset.
To achieve high precision results against the errors confused by the emotion classification we computes the distance as following procedure：1) If the query music file is classified to sad emotion class, the distance between the query signature and every signature with sad emotion subset in the music database is computed.2) The confusion set of the sad emotion subsets is the calm emotion subset. Thus, the distance between the query signature and every signature with the calm emotion subset in the music database is computed.

제안 방법

Therefore, three important factors are investigated to obtain the performance of the proposed system: fast feature extraction methods, efficient music emotion classification approach, and rapid on-line music similarity search scheme.
Generally, two songs belonging to quite different genre categories or bringing listeners (or users) different emotion are hardly thought as similar. Thus, in this paper we present an efficient algorithm to retrieve similar music files based on music emotion classification against a large archive of digital music.
Furthermore, considering the increasing improvements on portable devices, the system is designed to be compatible to the modern high-end MP3 players and smart phones. Therefore, three important factors are investigated to obtain the performance of the proposed system: fast feature extraction methods, efficient music emotion classification approach, and rapid on-line music similarity search scheme.
,24. In this paper, we use a Euclidean distance between two normalized vectors to measure the similarity of two songs on account of its computational economy. For p-th song and q-th song, the similarity measure is defined as (6).
The basic idea comes from the criterion： "two similar songs must be in the same music emotion". Using the acoustic-based distance measure between the pre-classification music file with the same type of emotion, the proposed method significantly speeds up the search process and increases the precision.
In this paper, we propose an efficient music similarity 응earch system based on the results of music emotion classification. A new tempo feature based on MDCT and wavelet transformation is combined with conventional timbre features in order to significantly improve the precision of music classification and similarity results. Furthermore, high precision of music classfication results in better recall rate and higher search speed than the traditional brute-force searching scheme.

대상 데이터

2000 songs are chosen as the testing set and labeled as 100 similar sets. Each similar set comprises 20 songs which have the same artist and the same emotion, and perceptually sound similar.

이론/모형

The related works on the topic of similarity query are presented in [1-7]. In [1], mel-frequecy cepstral coefficients(MFCCs) are adopted as the acoustic features and the distribution of MFCC is modeled by a Gaussian mixture model(GMM). Monte Carlo distance is used to calculate the distance between two songs based on their GMM.
The related works on the topic of similarity query are presented in [1-7]. In [1], mel-frequecy cepstral coefficients(MFCCs) are adopted as the acoustic features and the distribution of MFCC is modeled by a Gaussian mixture model(GMM). Monte Carlo distance is used to calculate the distance between two songs based on their GMM.

성능/효과

With AdaBoost algorithm combining timbre and tempo features, the music emotion classification precision can reach 93.2%. Using tempo features based on wavelet transform instead of fast Fourier transform we obtained the improved classification results although DB size gets larger compared to [8].
The distance between the pre-classified music files with the same type of emotion is smaller than the distance between the music files with the different type of emotion. With the emotion classification scheme, the searching speed, the average recall rate, and the average precision rate are significantly improved.

내보내기 구분	파일저장 인쇄 메일전송
구성항목	기본정보 상세정보 관리번호, 논문명, 저널/프로시딩명, 저자 , 발행년, 권, 호, 시작페이지, 끝페이지, 발행기관 관리번호, 논문명, 대등논문명, 저자 , 저널/프로시딩명, 발행기관, 발행년, 발행언어, 권, 호, 시작페이지, 끝페이지, ISBN, ISSN, 주제분야, 키워드, 초록(한글), 초록(영문), 저자(소속기관)
저장형식	Text(ASCII format) Excel format RefWorks Direct Export RIS format (for Reference Manager, ProCite, EndNote), Scholar's Aids, Mendeley
메일정보	받는사람 (필수) @ 보내는사람 (선택) @ 제목 내용 KISTI 검색결과 이메일 서비스
안내	총 건의 자료가 검색되었습니다. 다운받으실 자료의 인덱스를 입력하세요. (1-10,000) 검색결과의 순서대로 최대 10,000건 까지 다운로드가 가능합니다. 데이타가 많을 경우 속도가 느려질 수 있습니다.(최대 2~3분 소요) 다운로드 파일은 UTF-8 형태로 저장됩니다. 파일의 내용이 제대로 보이지 않을실 때는 웹브라우저 상단의 보기 -> 인코딩 -> 자동선택 여부를 확인하십시오. ~ Text(ASCII format) Excel format

연합인증

[국내논문] Music Similarity Search Based on Music Emotion Classification 원문보기

Abstract ▼ AI-Helper

Keyword

AI 본문요약
AI-Helper

문제 정의

가설 설정

제안 방법

대상 데이터

이론/모형

성능/효과

참고문헌 (8)

이 논문을 인용한 문헌

저자의 다른 논문 :

관련 콘텐츠

원문 보기

원문 URL 링크

오픈액세스(OA) 유형

이 논문과 함께 이용한 콘텐츠

AI-Helper ※ AI-Helper는 오픈소스 모델을 사용합니다.

선택된 텍스트

연합인증

[국내논문] Music Similarity Search Based on Music Emotion Classification 원문보기

Abstract ▼ AI-Helper

Keyword

AI 본문요약 엑셀 다운로드 AI-Helper

문제 정의

가설 설정

제안 방법

대상 데이터

이론/모형

성능/효과

참고문헌 (8)

이 논문을 인용한 문헌

저자의 다른 논문 :

김형국 (53)

관련 콘텐츠

원문 보기

원문 URL 링크

오픈액세스(OA) 유형

이 논문과 함께 이용한 콘텐츠

AI-Helper ※ AI-Helper는 오픈소스 모델을 사용합니다.

선택된 텍스트

AI 본문요약
AI-Helper