Arthur Flexer

A reliable probabilistic sleep stager based on a single EEG signal

OBJECTIVE We developed a probabilistic continuous sleep stager based on Hidden Markov models using only a single EEG signal. It offers the advantage of being objective by not relying on human scorers, having much finer temporal resolution (1s instead of 30s), and being based on solid probabilistic principles rather than a predefined set of rules (Rechtschaffen & Kales) METHODS AND MATERIAL Sixty-eight whole night sleep recordings from two different sleep labs are analysed using Gaussian observation Hidden Markov models. RESULTS Our unsupervised approach detects the cornerstones of human sleep (wakefulness, deep and REM sleep) with around 80% accuracy based on data from a single EEG channel. There are some difficulties in generalizing results across sleep labs. CONCLUSION Using data from a single electrode is sufficient for reliable continuous sleep staging. Sleep recordings from different sleep labs are not directly comparable. Training of separate models for the sleep labs is necessary.

The neglected user in music information retrieval research

Personalization and context-awareness are highly important topics in research on Intelligent Information Systems. In the fields of Music Information Retrieval (MIR) and Music Recommendation in particular, user-centric algorithms should ideally provide music that perfectly fits each individual listener in each imaginable situation and for each of her information or entertainment needs. Even though preliminary steps towards such systems have recently been presented at the “International Society for Music Information Retrieval Conference” (ISMIR) and at similar venues, this vision is still far away from becoming a reality. In this article, we investigate and discuss literature on the topic of user-centric music retrieval and reflect on why the breakthrough in this field has not been achieved yet. Given the different expertises of the authors, we shed light on why this topic is a particularly challenging one, taking computer science and psychology points of view. Whereas the computer science aspect centers on the problems of user modeling, machine learning, and evaluation, the psychological discussion is mainly concerned with proper experimental design and interpretation of the results of an experiment. We further present our ideas on aspects crucial to consider when elaborating user-aware music retrieval systems.

Using ICA for removal of ocular artifacts in EEG recorded from blind subjects

One of the standard applications of Independent Component Analysis (ICA) to EEG is removal of artifacts due to movements of the eye bulbs. Short blinks as well as slower saccadic movements are removed by subtracting respective independent components (ICs). EEG recorded from blind subjects poses special problems, since it shows a higher quantity of eye movements, which are also more irregular and very different across subjects. It is demonstrated that ICA can still be of use by comparing results from four blind subjects with results from one subject without eye bulbs who therefore does not show eye movement artifacts at all.

Statistical evaluation of music information retrieval experiments

Abstract This work concerns the necessity of statistical evaluation of Music Information Retrieval (MIR) experiments. This necessity is motivated by applying fundamental notions of statistical hypotheses testing to MIR research. Minimum requirements concerning statistical evaluation are developed and the appropriate statistical techniques are introduced and exemplified in a genre classification context. Articles from the MIR literature are examined and criticized for the lack of statistical evaluation they contain.

Hierarchical organization and description of music collections at the artist level

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.

Effects of album and artist filters in audio similarity computed for very large music databases

In music information retrieval, one of the central goals is to automatically recommend music to users based on a query song or query artist. This can be done using expert knowledge (e.g., www.pandora.com), social meta-data (e.g., www.last.fm), collaborative filtering (e.g., www.amazon.com/mp3), or by extracting information directly from the audio (e.g., www.muffin.com). In audio-based music recommendation, a wellknown effect is the dominance of songs from the same artist as the query song in recommendation lists. This effect has been studied mainly in the context of genre-classification experiments. Because no ground truth with respect to music similarity usually exists, genre classification is widely used for evaluation of music similarity. Each song is labelled as belonging to a music genre using, e.g., advice of a music expert. High genre classification results indicate good similarity measures. If, in genre classification experiments, songs from the same artist are allowed in both training and test sets, this can lead to over-optimistic results since usually all songs from an artist have the same genre label. It can be argued that in such a scenario one is doing artist classification rather than genre classification. One could even speculate that the specific sound of an album (mastering and production effects) is being classified. In Pampalk, Flexer, and Widmer (2005) the use of a so-called “artist filter” that ensures that a given artist’s songs are either all in the training set, or all in the test set, is proposed. Those authors found that the use of such an artist filter can lower the

A fast audio similarity retrieval method for millions of music tracks

We present a filter-and-refine method to speed up nearest neighbor searches with the Kullback–Leibler divergence for multivariate Gaussians. This combination of features and similarity estimation is of special interest in the field of automatic music recommendation as it is widely used to compute music similarity. However, the non-vectorial features and a non-metric divergence make using it with large corpora difficult, as standard indexing algorithms can not be used. This paper proposes a method for fast nearest neighbor retrieval in large databases which relies on the above approach. In its core the method rescales the divergence and uses a modified FastMap implementation to speed up nearest-neighbor queries. Overall the method accelerates the search for similar music pieces by a factor of 10–30 and yields high recall values of 95–99% compared to a standard linear search.

Limitations of interactive music recommendation based on audio content

We present a study on the limitations of an interactive music recommendation service based on automatic computation of audio similarity. Songs which are, according to the audio similarity function, similar to very many other songs and hence appear unwantedly often in recommendation lists keep a significant proportion of the audio collection from being recommended at all. This problem is studied in-depth with a series of computer experiments including analysis of alternative audio similarity functions and comparison with actual download data.

Identification of perceptual qualities in textural sounds using the repertory grid method

This paper is about exploring which perceptual qualities are relevant to people listening to textural sounds. Knowledge about those personal constructs shall eventually lead to more intuitive interfaces for browsing large sound libraries. By conducting mixed qualitative-quantitative interviews within the repertory grid framework ten bi-polar qualities are identified. A subsequent web-based study yields measures for inter-rater agreement and mutual similarity of the perceptual qualities based on a selection of 100 textural sounds. Additionally, some initial experiments are conducted to test standard audio descriptors for their correlation with the perceptual qualities.

Using hidden Markov models to build an automatic, continuous and probabilistic sleep stager

We report about an automatic continuous sleep stager which is based on probabilistic principles employing hidden Markov models (HMMs). Our sleep stager offers the advantage of being objective by not relying on human scorers, having much finer temporal resolution (1 second instead of 30 second): and being based on solid probabilistic principles rather than a predefined set of rules. Results obtained for nine whole night sleep recordings are reported.