• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06b
• /
• pp.405-408
• /
• 2011

기존의 작곡가, 곡명 등의 질의어로 대표될 수 있는 메타 데이터 기반 음악 검색은 음악 데이터의 양이 급증함에 따라 탐색 근거가 되는 사항을 모를 경우 사용자의 요구를 충족시켜 줄 수 없는 단점을 가지고 있다 [1]. 반면 음악의 내용을 기반으로 한 검색의 경우 이러한 제약에서 보다 자유로울 수 있다. 내용 기반 검색[2]의 연구에 있어서 음악 데이터로부터 사용자의 질의를 처리해주기 위한 멜로디를 추출해주는 것이 중요한 문제가 되는데, 데이터의 추출과 편집이 용이한 MIDI 파일이 많이 사용되고 있다. 본 논문에서는 사용자의 질의를 바탕으로 한 QBH system상에서의 음악 데이터 구축을 위해 MIDI파일에서 추출해 낼 수 있는 특징들을 이용해 MIDI파일의 멜로디 트랙 멜로디 트랙과 반주 트랙을 자동으로 구분하는 것을 목적으로, MIDI트랙에서 추출해서 이용할 수 있는 특징들에 대해 알아보고 그에 따른 간단한 분류 알고리즘의 제안과 실험 결과에 대해 소개한다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2013.06a
• /
• pp.381-384
• /
• 2013

본 논문에서는 허밍기반 음원 검색 시스템을 위한 오디오 특징 시퀀스 데이터 색인 기법을 제안한다. 우선 Query-by-Singing/Humming (QbSH) 시스템의 특징 데이터베이스를 생성하기 위하여 MP3 와 같은 다성음원에서 주요 멜로디를 추출하여 시퀀스데이터를 생성하고, 고속 검색을 지원하기 위한 시퀀스데이터를 색인화한다. 본 논문에서는 최소 Dynamic Time Warping (DTW) 거리 기법, 시퀀스 추상화 기법, 상한 값 기반 DTW 기법과 같이 세 가지의 시퀀스 데이터의 색인화 기술을 제시하고 각각에 대한 문제점을 파악하고, 성능을 평가한다. 이를 통하여 향상된 검색 시간과 검색 정확도를 얻을 수 있다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2012.07a
• /
• pp.257-259
• /
• 2012

이 논문에서는 실험을 통해서, 기존에 제안하였던 쿼리-바이-싱잉/허밍 (Query-by-singing/humming, QbSH) 시스템의 매칭 엔진의 강인성을 검증하고 그 결과를 제시한다. QbSH 시스템은 디지털 음악의 사용이 보편화되면서 음악 검색의 방법으로 많은 연구가 진행되어 오고 있다. QbSH 시스템은 입력으로부터 멜로디의 특징을 추출하는 부분과 추출된 특징을 매칭하는 부분으로 나눌 수 있는데, 매칭 단계에서 특징이 추출된 두 개의 멜로디 사이의 유사도 또는 거리를 계산하여 가장 유사한 멜로디를 데이터베이스에서 찾게 된다. 이 논문에서는 이 중, 기존에 제안하였던 매칭 엔진 부분의 강인성을 알아보기 위해서 입력으로부터 멜로디의 피치 시퀀스를 추출하는 과정을 간략히 하여 그 결과를 살펴보았다. 즉, 기존에 제안한 매칭 엔진이 특정한 피치 시퀀스 추출 과정에서만 좋은 성능을 보이는 게 아님을 실험을 통해서 살펴보았다. 실험 결과, 피치 시퀀스를 추출하는 과정이 극도로 간략해지더라도, 매칭 엔진을 좋은 성능을 보여주었다.

• Journal of Korea Multimedia Society
• /
• v.12 no.11
• /
• pp.1581-1592
• /
• 2009

Recently, we can hear various kinds of music in everywhere and anytime. If a user wants to find the music which was heard before in a street or cafe, but he does not know the title of the music, it is difficult to find it. That is the limitation of previous retrieval system of music data. To overcome these problems, we research a method of content-based retrieval of music data based on the recorded humming, the part of recorded music and the played musical instrument. In this paper, we investigated previous content-based retrieval methods of papers, systems and patents. Based on that, we research a method of content-based retrieval of music data. That is, in case of using the recorded humming and music for query, we extract the frequency information from the recorded humming/music and the stored music data by using FFT. We use a MIDI file in case of query by the played musical instrument. And by using dynamic programming matching, the error caused by the disparity of length between the input source with the stored music data could be reduced.

• Journal of Korea Multimedia Society
• /
• v.15 no.1
• /
• pp.18-31
• /
• 2012

This paper proposes a matching engine of query-by-singing/humming (QbSH) system which retrieves the most similar music information by comparing the input data with the extracted feature information from polyphonic music like MP3. The feature sequences transcribed from polyphonic music may have many errors. So, to reduce the influence of errors and improve the performance, the chroma-scale representation, compensation and asymmetric DTW (Dynamic Time Warping) are adopted in the matching engine. The performance of various distance metrics are also investigated in this paper. In our experiment, the proposed QbSH system achieves MRR (Mean Reciprocal Rank) of 0.718 for 1000 singing/humming queries when searching from a database of 450 polyphonic musics.

• Journal of KIISE:Databases
• /
• v.34 no.5
• /
• pp.409-419
• /
• 2007

Database community focuses on the similar music retrieval systems for music database when a humming query is given. One of the approaches is converting the midi data to time series, building their indices and performing the similarity search on them. Queries based on humming can be transformed to time series by using the known pitch detection algorithms. The recently suggested algorithm, scaled and warped matching, is based on dynamic time warping and uniform scaling. This paper proposes Humming BIRD(Humming Based sImilaR mini music retrieval system) using sliding window and center-aligned scaled and warped matching. Center-aligned scaled and warped matching is a mixed distance measure of center-aligned uniform scaling and time warping. The newly proposed measure gives tighter lower bound than previous ones which results in reduced search space. The empirical results show the superiority of this algorithm comparing the pruning power while it returns the same results.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.59 no.2
• /
• pp.445-451
• /
• 2010

This paper presents a threshold adaptation based voice query transcription scheme for music information retrieval. The proposed scheme analyzes monophonic voice signal and generates its transcription for diverse music retrieval applications. For accurate transcription, we propose several advanced features including (i) Energetic Feature eXtractor (EFX) for onset, peak, and transient area detection; (ii) Modified Windowed Average Energy (MWAE) for defining multiple small but coherent windows with local threshold values as offset detector; and finally (iii) Circular Average Magnitude Difference Function (CAMDF) for accurate acquisition of fundamental frequency (F0) of each frame. In order to evaluate the performance of our proposed scheme, we implemented a prototype music transcription system called AMT2 (Automatic Music Transcriber version 2) and carried out various experiments. In the experiment, we used QBSH corpus [1], adapted in MIREX 2006 contest data set. Experimental result shows that our proposed scheme can improve the transcription performance.

• The Journal of the Acoustical Society of Korea
• /
• v.23 no.6
• /
• pp.488-496
• /
• 2004

This paper describes a music information retrieval system which uses humming as the key for retrieval Humming is an easy way for the user to input a melody. However, there are several problems with humming that degrade the retrieval of information. One problem is a human factor. Sometimes people do not sing accurately, especially if they are inexperienced or unaccompanied. Another problem arises from signal processing. Therefore, a music information retrieval method should be sufficiently robust to surmount various humming errors and signal processing problems. A retrieval system has to extract pitch from the user's humming. However pitch extraction is not perfect. It often captures half or double pitches. even if the extraction algorithms take the continuity of the pitch into account. Considering these problems. we propose a system that takes multiple pitch candidates into account. In addition to the frequencies of the pitch candidates. the confidence measures obtained from their powers are taken into consideration as well. We also propose the use of an algorithm with three dimensions that is an extension of the conventional DP algorithm, so that multiple pitch candidates can be treated. Moreover in the proposed algorithm. DP paths are changed dynamically to take deltaPitches and IOIratios of input and reference notes into account in order to treat notes being split or unified. We carried out an evaluation experiment to compare the proposed system with a conventional system. From the experiment. the proposed method gave better retrieval performance than the conventional system.

• Journal of Broadcast Engineering
• /
• v.9 no.1
• /
• pp.2-7
• /
• 2004

In this paper, we proposed the content-based music information retrieval technique using some MPEG-7 low-level descriptors. Especially, pitch information and timbral features can be applied in music genre classification, music retrieval, or QBH(Query By Humming) because these can be modeling the stochasticpattern or timbral information of music signal. In this work, we restricted the music domain as O.S.T of movie or soap opera to apply broadcasting system. That is, the user can retrievalthe information of the unknown music using only an audio clip with a few seconds extracted from video content when background music sound greeted user's ear. We proposed the audio feature set organized by MPEG-7 descriptors and distance function by vector distance or ratio computation. Thus, we observed that the feature set organized by pitch information is superior to timbral spectral feature set and IFCR(Intra-Feature Component Ratio) is better than ED(Euclidean Distance) as a vector distance function. To evaluate music recognition, k-NN is used as a classifier

• Journal of the Korean Society for information Management
• /
• v.30 no.3
• /
• pp.23-48
• /
• 2013

This study compared and analyzed 15 CBMR (Content-based Music Retrieval) systems accessible on the web in terms of DB size and type, query type, access point, input and output type, and search functions, with reviewing features of music information and techniques used for transforming or transcribing of music sources, extracting and segmenting melodies, extracting and indexing features of music, and matching algorithms for CBMR systems. Application of text information retrieval techniques such as inverted indexing, N-gram indexing, Boolean search, truncation, keyword and phrase search, normalization, filtering, browsing, exact matching, similarity measure using edit distance, sorting, etc. to enhancing the CBMR; effort for increasing DB size and usability; and problems in extracting melodies, deleting stop notes in queries, and using solfege as pitch information were found as the results of analysis.