With Islands of Music we present a system which facilitates exploration of music libraries without requiring manual genre classification. Given pieces of music in raw audio format we estimate their perceived sound similarities based on psychoacoustic models. Subsequently, the pieces are organized on a 2-dimensional map so that similar pieces are located close to each other. A visualization using a metaphor of geographic maps provides an intuitive interface where islands resemble genres or styles of music. We demonstrate the approach using a collection of 359 pieces of music.

We present a study on the importance of psycho-acoustic transformations for effective audio feature calculation. From the results, both crucial and problematic parts of the algorithm for Rhythm Patterns feature extraction are identified. We furthermore introduce two new feature representations in this context: Statistical Spectrum Descriptors and Rhythm Histogram features. Evaluation on both the individual and combined feature sets is accomplished through a music genre classification task, involving 3 reference audio collections. Results are compared to published measures on the same data sets. Experiments confirmed that in all settings the inclusion of psycho-acoustic transformations provides significant improvement of classification accuracy.

With the rising popularity of digital music archives the need for new access methods such as interactive exploration or similarity-based search become significant. In this paper we present the PlaySOM, as well as the PocketSOMPlayer, two novel interfaces allowing to browse a music collection by navigating a map of clustered music tracks and to select regions of interest containing similar tracks for playing. The PlaySOM system is primarily designed to allow interaction via a large-screen device, whereas the PocketSOMPlayer is implemented for mobile devices, supporting both local as well as streamed audio replay. This approach offers content-based organization of music as an alternative to conventional navigation of audio archives, i.e. flat or hierarchical listings of music tracks that are sorted and filtered by meta information.

This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. Journal of New Music Research 0929-8215/03/3202-193$16.00 2003, Vol. 32, No. 2, pp. 193–210 © Swets & Zeitlinger

This survey paper provides an overview of content-based music information retrieval systems, both for audio and for symbolic music notation. Matching algorithms and indexing methods are briefly presented. The need for a TREC-like comparison of matching algorithms such as MIREX at ISMIR becomes clear from the high number of quite different methods which so far only have been used on different data collections. We placed the systems on a map showing the tasks and users for which they are suitable, and we find that existing content-based retrieval systems fail to cover a gap between the very general and the very specific retrieval tasks.

Abstract. As digital music collections grow, so does the need to organizing them automatically. In this paper we present an approach to hierarchically organize music collections at the artist level. Artists are grouped according to similarity which is computed using a web search engine and standard text retrieval techniques. The groups are described by words found on the webpages using term selection techniques and domain knowledge. We compare different term selection techniques, present a simple demonstration, and discuss our findings. 1

A software system that automatically classifies MIDI files into hierarchically organized taxonomies of musical genres is presented. This extensible software includes an easy to use and flexible GUI. An extensive library of high-level musical features is compiled, including many original features. A novel hybrid classification system is used that makes use of hierarchical, flat and round robin classification. Both k-nearest neighbour and neural network-based classifiers are used, and feature selection and weighting are performed using genetic algorithms. A thorough review of previous research in automatic genre classification is presented, along with an overview of automatic feature selection and classification techniques. Also included is a discussion of the theoretical issues relating to musical genre, including but not limited to what mechanisms humans use to classify music by genre and how realistic genre taxonomies can be constructed.

Multiple-instance learning algorithms train classifiers from lightly supervised data, i.e. labeled collections of items, rather than labeled items. We compare the multiple-instance learners mi-SVM and MILES on the task of classifying 10second song clips. These classifiers are trained on tags at the track, album, and artist levels, or granularities, that have been derived from tags at the clip granularity, allowing us to test the effectiveness of the learners at recovering the clip labeling in the training set and predicting the clip labeling for a held-out test set. We find that mi-SVM is better than a control at the recovery task on training clips, with an average classification accuracy as high as 87 % over 43 tags; on test clips, it is comparable to the control with an average classification accuracy of up to 68%. MILES performed adequately on the recovery task, but poorly on the test clips. 1

BeatBoxing is a type of vocal percussion, where musicians use their lips, cheeks, and throat to create different beats. It is commonly used by hiphop and rap artists. In this paper, we explore the use of BeatBoxing as a query mechanism for music information retrieval and more speci£cally the retrieval of drum loops. A classi£cation system that automatically identi£es the individual beat boxing sounds and can map them to corresponding drum sounds has been developed. In addition, the tempo of BeatBoxing is automatically detected and used to dynamically browse a database of music. We also describe some experiments in extracting structural representations of rhythm and their use for style classi£cation of drum loops. 1.

Abstract. Multimedia content can be described in versatile ways as its essence is not limited to one view. For music data these multiple views could be a song’s audio features as well as its lyrics. Both of these modalities have their advantages as text may be easier to search in and could cover more of the ‘content semantics ’ of a song, while omitting other types of semantic categorisation. (Psycho)acoustic feature sets, on the other hand, provide the means to identify tracks that ‘sound similar’ while less supporting other kinds of semantic categorisation. Those discerning characteristics of different feature sets meet users ’ differing information needs. We will explain the nature of text and audio feature sets which describe the same audio tracks. Moreover, we will propose the use of textual data on top of low level audio features for music genre classification. Further, we will show the impact of different combinations of audio features and textual features based on content words. 1