Exploring composite acoustic features for efficient music similarity query

Music similarity query based on acoustic content is becoming important with the ever-increasing growth of the music information from emerging applications such as digital libraries and WWW. However, relative techniques are still in their infancy and much less than satisfactory. In this paper, we present a novel index structure, called Composite Feature tree, CF-tree, to facilitate efficient content-based music search adopting multiple musical features. Before constructing the tree structure, we use PCA to transform the extracted features into a new space sorted by the importance of acoustic features. The CF-tree is a balanced multi-way tree structure where each level represents the data space at different dimensionalities. The PCA transformed data and reduced dimensions in the upper levels can alleviate suffering from dimensionality curse. To accurately mimic human perception, an extension, named CF+-tree, is proposed, which further applies multivariable regression to determine the weight of each individual feature. We conduct extensive experiments to evaluate the proposed structures against state-of-art techniques. The experimental results demonstrate superiority of our technique.

Machine learning techniques for music information retrieval

Tese de doutoramento, Informática (Engenharia Informática), Universidade de Lisboa, Faculdade de Ciências, 2015

Algorithmes de recommandation musicale

Ce mémoire est composé de trois articles qui s’unissent sous le thème de la recommandation musicale à grande échelle. Nous présentons d’abord une méthode pour effectuer des recommandations musicales en récoltant des étiquettes (tags) décrivant les items et en utilisant cette aura textuelle pour déterminer leur similarité. En plus d’effectuer des recommandations qui sont transparentes et personnalisables, notre méthode, basée sur le contenu, n’est pas victime des problèmes dont souffrent les systèmes de filtrage collaboratif, comme le problème du démarrage à froid (cold start problem). Nous présentons ensuite un algorithme d’apprentissage automatique qui applique des étiquettes à des chansons à partir d’attributs extraits de leur fichier audio. L’ensemble de données que nous utilisons est construit à partir d’une très grande quantité de données sociales provenant du site Last.fm. Nous présentons finalement un algorithme de génération automatique de liste d’écoute personnalisable qui apprend un espace de similarité musical à partir d’attributs audio extraits de chansons jouées dans des listes d’écoute de stations de radio commerciale. En plus d’utiliser cet espace de similarité, notre système prend aussi en compte un nuage d’étiquettes que l’utilisateur est en mesure de manipuler, ce qui lui permet de décrire de manière abstraite la sorte de musique qu’il désire écouter.

Apprentissage statistique pour l'étiquetage de musique et la recommandation

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal

TraSe algorithm : automatic evolutionary morphing of electronic dance music

The TraSe (Transform-Select) algorithm has been developed to investigate the morphing of electronic music through automatically applying a series of deterministic compositional transformations to the source, guided towards a target by similarity metrics. This is in contrast to other morphing techniques such as interpolation or parameters or probabilistic variation. TraSe allows control over stylistic elements of the music through user-defined weighting of numerous compositional transformations. The formal evaluation of TraSe was mostly qualitative and occurred through nine participants completing an online questionnaire. The music generated by TraSe was generally felt to be less coherent than a human composed benchmark but in some cases judged as more creative.

Note sequence morphing algorithms for performance of electronic dance music

This paper describes algorithms that can musically augment the realtime performance of electronic dance music by generating new musical material by morphing. Note sequence morphing involves the algorithmic generation of music that smoothly transitions between two existing musical segments. The potential of musical morphing in electronic dance music is outlined and previous research is summarised; including discussions of relevant music theoretic and algorithmic concepts. An outline and explanation is provided of a novel Markov morphing process that uses similarity measures to construct transition matrices. The paper reports on a ‘focus-concert’ study used to evaluate this morphing algorithm and to compare its output with performances from a professional DJ. Discussions of this trial include reflections on some of the aesthetic characteristics of note sequence morphing. The research suggests that the proposed morphing technique could be effectively used in some electronic dance music contexts.

On Musical Self-Similarity : Intersemiosis as Synecdoche and Analogy

Self-similarity, a concept taken from mathematics, is gradually becoming a keyword in musicology. Although a polysemic term, self-similarity often refers to the multi-scalar feature repetition in a set of relationships, and it is commonly valued as an indication for musical coherence and consistency . This investigation provides a theory of musical meaning formation in the context of intersemiosis, that is, the translation of meaning from one cognitive domain to another cognitive domain (e.g. from mathematics to music, or to speech or graphic forms). From this perspective, the degree of coherence of a musical system relies on a synecdochic intersemiosis: a system of related signs within other comparable and correlated systems. This research analyzes the modalities of such correlations, exploring their general and particular traits, and their operational bounds. Looking forward in this direction, the notion of analogy is used as a rich concept through its two definitions quoted by the Classical literature: proportion and paradigm, enormously valuable in establishing measurement, likeness and affinity criteria. Using quantitative qualitative methods, evidence is presented to justify a parallel study of different modalities of musical self-similarity. For this purpose, original arguments by Benoît B. Mandelbrot are revised, alongside a systematic critique of the literature on the subject. Furthermore, connecting Charles S. Peirce s synechism with Mandelbrot s fractality is one of the main developments of the present study. This study provides elements for explaining Bolognesi s (1983) conjecture, that states that the most primitive, intuitive and basic musical device is self-reference, extending its functions and operations to self-similar surfaces. In this sense, this research suggests that, with various modalities of self-similarity, synecdochic intersemiosis acts as system of systems in coordination with greater or lesser development of structural consistency, and with a greater or lesser contextual dependence.

Query Click and Text Similarity Graph for Query Suggestions

Query suggestion is an important feature of the search engine with the explosive and diverse growth of web contents. Different kind of suggestions like query, image, movies, music and book etc. are used every day. Various types of data sources are used for the suggestions. If we model the data into various kinds of graphs then we can build a general method for any suggestions. In this paper, we have proposed a general method for query suggestion by combining two graphs: (1) query click graph which captures the relationship between queries frequently clicked on common URLs and (2) query text similarity graph which finds the similarity between two queries using Jaccard similarity. The proposed method provides literally as well as semantically relevant queries for users' need. Simulation results show that the proposed algorithm outperforms heat diffusion method by providing more number of relevant queries. It can be used for recommendation tasks like query, image, and product suggestion.

Effects of timbre and tempo change on memory for music

We investigated the effects of different encoding tasks and of manipulations of two supposedly surface parameters of music on implicit and explicit memory for tunes. In two experiments, participants were first asked to either categorize instrument or judge familiarity of 40 unfamiliar short tunes. Subsequently, participants were asked to give explicit and implicit memory ratings for a list of 80 tunes, which included 40 previously heard. Half of the 40 previously heard tunes differed in timbre (Experiment 1) or tempo (Experiment 2) in comparison with the first exposure. A third experiment compared similarity ratings of the tunes that varied in timbre or tempo. Analysis of variance (ANOVA) results suggest first that the encoding task made no difference for either memory mode. Secondly, timbre and tempo change both impaired explicit memory, whereas tempo change additionally made implicit tune recognition worse. Results are discussed in the context of implicit memory for nonsemantic materials and the possible differences in timbre and tempo in musical representations.

Perception of emotion in sounded and imagined music

WE STUDIED THE EMOTIONAL RESPONSES BY MUSICIANS to familiar classical music excerpts both when the music was sounded, and when it was imagined.We used continuous response methodology to record response profiles for the dimensions of valence and arousal simultaneously and then on the single dimension of emotionality. The response profiles were compared using cross-correlation analysis, and an analysis of responses to musical feature turning points, which isolate instances of change in musical features thought to influence valence and arousal responses. We found strong similarity between the use of an emotionality arousal scale across the stimuli, regardless of condition (imagined or sounded). A majority of participants were able to create emotional response profiles while imagining the music, which were similar in timing to the response profiles created while listening to the sounded music.We conclude that similar mechanisms may be involved in the processing of emotion in music when the music is sounded and when imagined.

User-centric music information retrieval

The rapid growth of the Internet and the advancements of the Web technologies have made it possible for users to have access to large amounts of on-line music data, including music acoustic signals, lyrics, style/mood labels, and user-assigned tags. The progress has made music listening more fun, but has raised an issue of how to organize this data, and more generally, how computer programs can assist users in their music experience. An important subject in computer-aided music listening is music retrieval, i.e., the issue of efficiently helping users in locating the music they are looking for. Traditionally, songs were organized in a hierarchical structure such as genre->artist->album->track, to facilitate the users’ navigation. However, the intentions of the users are often hard to be captured in such a simply organized structure. The users may want to listen to music of a particular mood, style or topic; and/or any songs similar to some given music samples. This motivated us to work on user-centric music retrieval system to improve users’ satisfaction with the system. The traditional music information retrieval research was mainly concerned with classification, clustering, identification, and similarity search of acoustic data of music by way of feature extraction algorithms and machine learning techniques. More recently the music information retrieval research has focused on utilizing other types of data, such as lyrics, user-access patterns, and user-defined tags, and on targeting non-genre categories for classification, such as mood labels and styles. This dissertation focused on investigating and developing effective data mining techniques for (1) organizing and annotating music data with styles, moods and user-assigned tags; (2) performing effective analysis of music data with features from diverse information sources; and (3) recommending music songs to the users utilizing both content features and user access patterns.