Audio Cover Song Identification and Similarity: Background, Approaches, Evaluation, and Beyond

Serrà, Joan; Herrera, Perfecto

doi:10.1007/978-3-642-11674-2_14

Cited by 90 publications

(108 citation statements)

References 46 publications

(135 reference statements)

Supporting

Mentioning

100

Contrasting

Unclassified

Order By: Relevance

“…Yu et al [4] also use LSH to compare song statistics. An overview of methods for cover song recognition can be found in [5]. For a given query song, our method prunes a database with a combination of rejectors that returns a small subset of the original dataset.…”

Section: Introductionmentioning

confidence: 99%

Efficient database pruning for large-scale cover song recognition

Osmalskyj

Pierard

Droogenbroeck

et al. 2013

2013 IEEE International Conference on Acoustics, Speech and Signal Processing

View full text Add to dashboard Cite

This paper focuses on cover song recognition over a large dataset, potentially containing millions of songs. At this time, the problem of cover song recognition is still challenging and only few methods have been proposed on large scale databases. We present an efficient method for quickly extracting a small subset from a large database in which a correspondence to an audio query should be found. We make use of fast rejectors based on independent audio features. Our method mixes independent rejectors together to build composite ones. We evaluate our system with the Million Song Dataset and we present composite rejectors offering a good trade-off between the percentage of pruning and the percentage of loss.

show abstract

Section: Introductionmentioning

confidence: 99%

Efficient database pruning for large-scale cover song recognition

Osmalskyj

Pierard

Droogenbroeck

et al. 2013

2013 IEEE International Conference on Acoustics, Speech and Signal Processing

View full text Add to dashboard Cite

show abstract

“…In addition, research on version identification can yield valuable clues on how music similarity can be modeled. This task, of automatically detecting versions of the same musical piece, has received much attention from the research community over recent years (see [36] for a survey). Potential applications range from the detection of copyright violations on websites such as YouTube, to the automation of computational analyses of musical influence networks [4].…”

Section: Introductionmentioning

confidence: 99%

“…Systems for the automatic detection of versions exploit musical facets which remain mostly unchanged across different renditions, primarily tonal information [36]. In this context, the term tonality refers, in a broad sense, to "the systematic arrangement of pitch phenomena and relations between them" [16].…”

Section: Introductionmentioning

confidence: 99%

“…As such, chroma features are related to harmony, understood as "the combining of notes simultaneously, to produce chords, and successively, to produce chord progressions" [5]. Common techniques for comparing sequences of chroma features include dynamic time warping and simple cross-correlation [36]. Another tonal representation that has been considered for version identification is the predominant melody, either by attempting to fully transcribe it [42], or by using it as a mid-level representation for computing similarity [23].…”

Section: Introductionmentioning

confidence: 99%

“…Whilst good results have been achieved using single music representations (in particular harmony represented by chroma features [36]), some recent studies suggest that version detection could be improved through the combination of different musical cues [9,19,35]. However, not much research has been carried out in this direction.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Tonal representations for music retrieval: from version identification to query-by-humming

Salamon

Serrà

Gómez

2012

Int J Multimed Info Retr

Self Cite

View full text Add to dashboard Cite

In this study we compare the use of different music representations for retrieving alternative performances of the same musical piece, a task commonly referred to as version identification. Given the audio signal of a song, we compute descriptors representing its melody, bass line and harmonic progression using state-of-the-art algorithms. These descriptors are then employed to retrieve different versions of the same musical piece using a dynamic programming algorithm based on nonlinear time series analysis. First, we evaluate the accuracy obtained using individual descriptors, and then we examine whether performance can be improved by combining these music representations (i.e. descriptor fusion). Our results show that whilst harmony is the most reliable music representation for version identification, the melody and bass line representations also carry useful information for this task. Furthermore, we show that by combining these tonal representations we can increase version detection accuracy. Finally, we demonstrate how the proposed version identification method can be adapted for the task of query-by-humming. We propose a melody-based retrieval approach, and demonstrate how melody representations extracted from recordings of a cappella singing can be successfully used to retrieve the original song from a collection of polyphonic audio. The current limitations of the proposed approach are discussed in the context of version identification and query-by-humming, and possible solutions and future research directions are proposed.

show abstract

Audio Cover Song Identification and Similarity: Background, Approaches, Evaluation, and Beyond

Cited by 90 publications

References 46 publications

Efficient database pruning for large-scale cover song recognition

Efficient database pruning for large-scale cover song recognition

Tonal representations for music retrieval: from version identification to query-by-humming

Contact Info

Product

Resources

About