Genre classification using chords and stochastic language models

Pérez-Sancho, Carlos; Rizo, David; Iñesta, José M.

doi:10.1080/09540090902733780

Cited by 34 publications

(27 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since the amount of digitized music has increased enormously over the past years, the need for tools to organize, order and cluster music has increased as well. Automatic music classification has already been investigated on the basis of many different ideas, among others compression distance [2], Hidden Markov Models [1], stochastic language models [12], and n-gram models [5].…”

Section: Introductionmentioning

confidence: 99%

Clustering and Classification of Music by Interval Categories

Honingh

Bod

2011

Mathematics and Computation in Music

View full text Add to dashboard Cite

Abstract. In this paper, we present a novel approach to clustering and classification of music. The approach is based on the concept of interval categories, which is a theoretical, but perceptually motivated construct that groups sets of pitch events into six different categories each of which has its own musical character. A piece of music can be represented by six percentages that express the occurrences of each category in the piece. This piece of music can, in this way, be represented as one point in a six dimensional space. We choose the three most significant dimensions to visualize music from different composers and genres. We will see that, using this approach, automatic classification between tonal and atonal music is possible. Furthermore, we will present a successful visual clustering of 1) the three periods of Beethoven, 2) the genres Rock and Jazz, and 3) several composers through various musical time periods.

show abstract

Section: Introductionmentioning

confidence: 99%

Clustering and Classification of Music by Interval Categories

Honingh

Bod

2011

Mathematics and Computation in Music

View full text Add to dashboard Cite

show abstract

“…[26] examined the statistics of the chord sequences of several thousand songs, and compared the results to those from a standard natural language corpus in an attempt to find lexical units in harmony that correspond to words in language. [34,35] investigated whether stochastic language models including naive Bayes classifiers and 2-, 3-and 4-grams could be used for automatic genre classification. The models were tested on both symbolic and audio data, where an off-the-shelf chord transcription algorithm was used to convert the audio data to a symbolic representation.…”

Section: Introductionmentioning

confidence: 99%

Probabilistic and Logic-Based Modelling of Harmony

Dixon

Mauch

Anglade

2011

Exploring Music Contents

View full text Add to dashboard Cite

Abstract. Many computational models of music fail to capture essential aspects of the high-level musical structure and context, and this limits their usefulness, particularly for musically informed users. We describe two recent approaches to modelling musical harmony, using a probabilistic and a logic-based framework respectively, which attempt to reduce the gap between computational models and human understanding of music. The first is a chord transcription system which uses a high-level model of musical context in which chord, key, metrical position, bass note, chroma features and repetition structure are integrated in a Bayesian framework, achieving state-of-the-art performance. The second approach uses inductive logic programming to learn logical descriptions of harmonic sequences which characterise particular styles or genres. Each approach brings us one step closer to modelling music in the way it is conceptualised by musicians.

show abstract

“…In this paper, a different approach is proposed to overcome those drawbacks, using a general-purpose encoding method for melodies, while music analysis is performed using language modeling, a technique that has proven to be very effective in music classification [2,8].…”

Section: Previous Work On Modeling Composer Stylementioning

confidence: 99%

“…A thorough review on the different uses of musical style in MIR tasks can be found in [9]. In a previous work [8] we used language modeling techniques to model the styles of different musical works in order to recognize their composers, with successful results. In this work we will focus on modeling the style of different composers.…”

Section: Introductionmentioning

confidence: 99%

Modeling Musical Style with Language Models for Composer Recognition

Hontanilla

Pérez-Sancho

Iñesta

2013

Pattern Recognition and Image Analysis

Self Cite

View full text Add to dashboard Cite

Abstract. In this paper we present an application of language modeling using n-grams to model the style of different composers. For this, we repeated the experiments performed in previous works by other authors using a corpus of 5 composers from the Baroque and Classical periods. In these experiments we found some signs that the results could be influenced by external factors other than the composers' styles, such as the heterogeneity in the musical forms selected for the corpus. In order to assess the validity of the modeling techniques to capture the own personal style of the composers, a new experiment was performed with a corpus of fugues from Bach and Shostakovich. All these experiments show that language modeling is a suitable tool for modeling musical style, even when the styles of the different datasets are affected by several factors.

show abstract

Genre classification using chords and stochastic language models

Cited by 34 publications

References 9 publications

Clustering and Classification of Music by Interval Categories

Clustering and Classification of Music by Interval Categories

Probabilistic and Logic-Based Modelling of Harmony

Modeling Musical Style with Language Models for Composer Recognition

Contact Info

Product

Resources

About