Matthias Mauch

• PhD
• Queen Mary University of London

Most of the literature on music recommendation assumes that a library of songs is available, and that the songs are neatly deduplicated and organ-ised into recommendable items. Significant effort has to go into even coming close to this ideal.

Here we investigate the evolutionary dynamics of several kinds of modern cultural artefacts—pop music, novels, the clinical literature and cars—as well as a collection of organic populations. In contrast to the general belief that modern culture evolves very quickly, we show that rates of modern cultural evolution are comparable to those of many an...

Sometimes the normal course of events is disrupted by a particularly swift and profound change. Historians have often referred to such changes as “revolutions”, and, though they have identified many of them, they have rarely supported their claims with statistical evidence. Here, we present a method to identify revolutions based on a measure of mul...

In musicology, there has been a long debate about a meaningful partitioning and description of music history regarding composition styles. Particularly, concepts of historical periods have been criticized since they cannot account for the continuous and interwoven evolution of style. To systematically study this evolution, large corpora are necessa...

This paper presents a feedback framework that can improve chord recognition for music audio signals by performing approximate note transcription with Bayesian non-negative matrix factorization (NMF) using prior knowledge on chords. Although the names and note compositions of chords are intrinsically linked with each other (e.g., C major chords are...

We investigate piano acoustics and compare the theoretical temporal decay of individual partials to recordings of real-world piano notes from the RWC Music Database. We first describe the theory behind double decay and beats, known phenomena caused by the interaction between strings and soundboard. Then we fit the decay of the first 30 partials to...

In this paper we improve piano note tracking using a Hid den Markov Model (HMM). We first transcribe piano music based on a non-negative matrix factorisation (NMF) method. For each note four templates are trained to represent the dif ferent stages of piano sounds: silence, attack, decay and re lease. Then a four-state HMM is employed to track notes...

Music, like language and genes, is the product of a descent-by-modification process. As such, the current distribution of music styles around the world should reflect the history of human migration and cultural diffusion (Lomax, 1968). However, where geneticists and linguists have developed sophisticated techniques for reconstructing that history,...

We present Tony, a software tool for the interactive annotation of melodies from monophonic audio recordings, and evaluate its usability and the accuracy of its note extraction method. The scientific study of acoustic performances of melodies, whether sung or played, requires the accurate transcription of notes and pitches. To achieve the desired t...

In modern societies, cultural change seems ceaseless. The flux of fashion is especially obvious for popular music. While much has been written about the origin and evolution of pop, most claims about its history are anecdotal rather than scientific in nature. To rectify this we investigate the US Billboard Hot 100 between 1960 and 2010. Using Music...

This paper presents a comparative study of classification performance in automatic audio chord recognition based on three chroma feature implementations, with the aim of distinguishing effects of frame size, instrumentation, and choice of chroma feature. Until recently, research in automatic chord recognition has focused on the development of compl...

We introduce MedleyDB: a dataset of annotated, royalty-free multitrack recordings. The dataset was primarily developed to support research on melody extraction, address-ing important shortcomings of existing collections. For each song we provide melody f 0 annotations as well as instrument activations for evaluating automatic instrument recognition...

This paper presents a study on intonation and intonation drift in unaccompanied singing, and proposes a simple model of reference pitch memory that accounts for many of the effects observed. Singing experiments were conducted with 24 singers of varying ability under three conditions (Normal, Masked, Imagined). Over the duration of a recording, ∼50...

We propose the Probabilistic YIN (PYIN) algorithm, a modification of the well-known YIN algorithm for fundamental frequency (F0) estimation. Conventional YIN is a simple yet effective method for frame-wise monophonic F0 estimation and remains one of the most popular methods in this domain. In order to eliminate short-term errors, outputs of frequen...

We propose string compressibility as a descriptor of temporal structure in audio, for the purpose of determining musical similarity. Our descriptors are based on computing track-wise compression rates of quantised audio features, using multiple temporal resolutions and quantisation granularities. To verify that our descriptors capture musically rel...

We showed that it is possible to produce music by a Darwinian process in which consumers select particular tunes over others (1). Our objective was to isolate the role of one element of the evolutionary process: consumer selection. In our system, heritable variation was generated by a computer algorithm. Claidiere et al. (2) argue that our methodol...

In this keynote talk, we describe two crowdsourcing-based web services, PodCastle (http://en.podcastle.jp for the English version and http://podcastle.jp for the Japanese version) and Songle (http://songle.jp). PodCastle and Songle collect voluntary contributions by anonymous users in order to improve the experiences of users listening to speech an...

In this keynote talk, we describe two crowdsourcing-based web services, PodCastle (http://en.podcastle.jp for the English version and http://podcastle.jp for the Japanese version) and Songle (http://songle.jp). PodCastle and Songle collect voluntary contributions by anonymous users in order to improve the experiences of users listening to speech an...

Music evolves as composers, performers, and consumers favor some musical variants over others. To investigate the role of consumer selection, we constructed a Darwinian music engine consisting of a population of short audio loops that sexually reproduce and mutate. This population evolved for 2,513 generations under the selective influence of 6,931...

Aligning lyrics to audio has a wide range of applications such as the automatic generation of karaoke scores, song-browsing by lyrics, and the generation of audio thumbnails. Existing methods are restricted to using only lyrics and match them to phoneme features extracted from the audio (usually mel-frequency cepstral coefficients). Our novel idea...

The inharmonicity of vibrating strings can easily be estimated from recordings of isolated tones. Likewise, the tuning system (temperament) of a keyboard instrument can be ascertained from isolated tones by estimating the fundamental frequencies corresponding to each key of the instrument. This paper addresses a more difficult problem: the automati...

This paper describes two web services, PodCastle and Songle, that collect voluntary contributions by anonymous users in order to im-prove the experiences of users listening to speech and music con-tent available on the web. These services use automatic speech-recognition and music-understanding technologies to provide con-tent analysis results, suc...

We present a corpus-based study of musical rhythm, based on a collection of 4.8 million bar-length drum patterns extracted from 48,176 pieces of symbolic music. Ap-proaches to the analysis of rhythm in music information retrieval to date have focussed on low-level features for re-trieval or on the detection of tempo, beats and drums in audio record...

This paper describes a public web service for active music listening, Songle, that enriches music listening experiences by using music-understanding technologies based on signal processing. Although various research-level interfaces and technologies have been developed, it has not been easy to get people to use them in everyday life. Songle serves...

We propose the novel audio feature structural change for the analysis and visualisation of recorded music, and argue that it is related to a particular notion of musical complexity. Structural change is a meta feature that can be calculated from an arbitrary frame-wise basis feature, with each element in the structural change feature vector represe...

Issues concerning tuning and temperament bear relevance to music research in areas such as historical musicology, performance and recording studies, and music perception. We have recently demonstrated that it is possible to classify keyboard temperament automatically from audio recordings of standard musical works to the extent of accurately distin...

We present a new genre classification framework using both low-level signal-based features and high-level harmony features. A state-of-the-art statistical genre classifier based on timbral features is extended using a first-order random forest containing for each genre rules derived from harmony or chord sequences. This random forest has been autom...

Chord labels provide a concise description of musical harmony. In pop and jazz music, a sequence of chord labels is often the only written record of a song, and forms the basis of so-called lead sheets. We devise a fully automatic method to simultaneously estimate from an audio waveform the chord sequence including bass notes, the metric positions...

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...

We present a novel music signal processing task of classifying the tuning of a harpsichord from audio recordings of standard musical works. We report the results of a classification experiment involving six different temperaments, using real harpsichord recordings as well as synthesised audio data. We introduce the concept of conservative transcrip...

In collections of recordings of classical music, it is normal to find multiple performances, usually by different artists, of the same pieces of music. While there may be differences in many dimensions of musical similarity, such as timbre, pitch or structural detail, the underlying musical content is essentially and recognizably the same. The degr...

The automatic detection and transcription of musical chords from audio is an established music computing task. The choice of chord profiles and higher-level time-series modelling have received a lot of attention, resulting in methods with an overall performance of more than 70% in the MIREX Chord Detection task 2009. Research on the front end of ch...

We propose two novel lyrics-to-audio alignment methods which make use of additional chord information. In the first method we extend an existing hidden Markov model (HMM) for lyrics alignment [1] by adding a chord model based on the chroma features often used in automatic audio chord detection. However, the textual transcriptions found on the Inter...

Chord extraction from audio is a well-established music computing task, and many valid approaches have been pre- sented in recent years that use different chord templates, smoothing techniques and musical context models. The present work shows that additional exploitation of the repet- itive structure of songs can enhance chord extraction, by combi...

This paper describes our approach to chord extraction from audio, a variant of which was submitted to the 2009 MIREX Chord Detection Task (No Training), and achieved the top ranking of 71.2%. The structural segmentation algorithm is a pre-processing step for the chord extraction, and was also submitted separately for the Structural Segmentation Tas...

Chord extraction from audio is a well-established music computing task, and many valid approaches have been presented in recent years that use different chord templates, smoothing techniques and musical context models. The present work shows that additional exploitation of the repetitive structure of songs can enhance chord extraction, by combining...

Chord labels for recorded audio are in high demand both as an end product used by musicologists and hobby musi- cians and as an input feature for music similarity applica- tions. Many past algorithms for chord labelling are based on chromagrams, but distribution of energy in chroma frames is not well understood. Furthermore, non-chord notes com- pl...

The availability of large, electronically encoded text cor-pora and the use of computers in recent decades have made Natural Language Processing (NLP) a flourishing research area. A wealth of standard techniques has been developed to serve use cases like document retrieval, identification of a finite vocabulary and synonyms, and the collocation of...

ABSTRACT Modern collections of symbolic and audio music content provide unprecedented,possibilities for musicological re- search, but traditional qualitative evaluation methods can- not realistically cope with such amounts,of data. We are interested in harmonic analysis and propose key-independ- ent chord idioms derived from a bottom-up analysis of...

We present our submission MD1 to the Chord Estima-tion Task of the 2010 Music Information Retrieval Eval-uation eXchange (MIREX 2010). The front-end chroma generation is based on a new implementation of NNLS Chroma 1 as a Vamp plugin. The higher level model is implemented in Matlab as a dynamic Bayesian network (DBN) with extensive context modellin...