REpeating Pattern Extraction Technique (REPET): A Simple Method for Music/Voice Separation

Rafii, Zafar; Pardo, Bryan

doi:10.1109/tasl.2012.2213249

Cited by 172 publications

(82 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…It is a topic that has many applications in the entertainment industry such as automatic karaoke [19], [29], music upmixing [21], [22], [23] or audio restoration [31]. For this reason, it has gathered the attention of a large community of researchers in the past 15 years [35], [34].…”

Section: Introductionmentioning

confidence: 99%

“…For instance, if we know that a musical background is repetitive whereas the vocal signal is not, it is much more efficient to enforce this knowledge rather than to choose a NMF model. This line of thought leads to the REPET algorithm [28], [19], [29], that proved very efficient for music/voice separation. Likewise, if our objective is to separate harmonic and percussive sounds, there is no real advantage in trying to build dictionaries of such sounds to use for NMF as in [15].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Scalable audio separation with light Kernel Additive Modelling

Liutkus

Fitzgerald

Rafii

2015

2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

Recently, Kernel Additive Modelling (KAM) was proposed as a unified framework to achieve multichannel audio source separation. Its main feature is to use kernel models for locally describing the spectrograms of the sources. Such kernels can capture source features such as repetitivity, stability over time and/or frequency, self-similarity, etc. KAM notably subsumes many popular and effective methods from the state of the art, including REPET and harmonic/percussive separation with median filters. However, it also comes with an important drawback in its initial form: its memory usage badly scales with the number of sources. Indeed, KAM requires the storage of the full-resolution spectrogram for each source, which may become prohibitive for full-length tracks or many sources. In this paper, we show how it can be combined with a fast compression algorithm of its parameters to address the scalability issue, thus enabling its use on small platforms or mobile devices.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Scalable audio separation with light Kernel Additive Modelling

Liutkus

Fitzgerald

Rafii

2015

2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

show abstract

“…Music accompaniment in a low-rank subspace. Repetition of music is a main parameter in a song [16], [19]. Singing voice is relatively sparse due to its variations or different pitch ranges within the songs.…”

Section: Introductionmentioning

confidence: 99%

Separation of Singing Voice from Music Background

Burute¹,

Mane²

2015

IJCA

View full text Add to dashboard Cite

Songs are representation of audio signal and musical instruments. An audio signal separation system should be able to identify different audio signals such as speech, background noise and music. In a song the singing voice provides useful information regarding pitch range, music content, music tempo and rhythm. An automatic singing voice separation system is used for attenuating or removing the music accompaniment. The paper presents survey of the various algorithm and method for separating singing voice from musical background. From the survey it is observed that most of researchers used Robust Principal Component Analysis method for separation of singing voice from music background, by taking into account the rank of music accompaniment and the sparsity of singing voices.

show abstract

“…For example, vocalized imitations [8] and artificially synthesized approximations [9] of the source of interest have been used as priors to improve separation results. Similarly, exactly [10] and approximately [11] repeated patterns in a piece of music have been used to improve the extraction of the remaining varying components.…”

Section: Introductionmentioning

confidence: 99%

Estimating timing and channel distortion across related signals

Raffel

Ellis

2014

2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

We consider the situation where there are multiple audio signals whose relationship is of interest. If these signals have been differently captured, the otherwise similar signals may be distorted by fixed filtering and/or unsynchronized timebases. Examples include recordings of signals before and after radio transmission and different versions of musical mixes obtained from CDs and vinyl LPs. We present techniques for estimating and correcting timing and channel differences across related signals. Our approach is evaluated in the context of artificially manipulated speech utterances and two source separation tasks.

show abstract

REpeating Pattern Extraction Technique (REPET): A Simple Method for Music/Voice Separation

Cited by 172 publications

References 16 publications

Scalable audio separation with light Kernel Additive Modelling

Scalable audio separation with light Kernel Additive Modelling

Separation of Singing Voice from Music Background

Estimating timing and channel distortion across related signals

Contact Info

Product

Resources

About