BibTeX:
@proceedings{lsas2009,,
  title = {{Proceedings of the 3rd Workshop on Learning the Semantics of Audio Signals (LSAS)}},
  editor = {Stephan Baumann and Juan Jos\'{e} Burred and Andreas N\"{u}rnberger and Sebastian Stober},
  address = {Graz, Austria},
  month = {Dec},
  year = {2009},
  isbn = {978-3-940961-38-9},
  url = {http://lsas2009.dke-research.de/proceedings/lsas2009proceedings.pdf}
}
Abstract: Music signals are highly structured data items where different elements combine at various levels of abstraction to create the desired result. This structure is not appropriately taken into account in conventional signal analysis methods, where the overall signal is characterized by calculating straightforward statistical measures in successive time frames. This talk introduces methods for breaking up a complex music signal into its constituent musical elements that have more well-defined "semantic" roles than the entire mixture signal. Methods are discussed for analyzing the vocals and lyrics of music pieces, extracting the melody, the bass line, and chords from music, recognizing musical instruments in complex music, and analyzing the rhythm and sectional form of music. Particular emphasis is placed on novel end-user applications that are enabled by these advanced signal analysis approaches. The applications include new interfaces and techniques for music information retrieval, intelligent music processing tools, and informative music playback interfaces where links to other music pieces are shown at localized segments of the played piece. Techniques for implementing these applications are discussed.
BibTeX:
@inproceedings{lsas2009klapuri,
  author = {Anssi Klapuri},
  title = {Extracting meaningful auditory objects from music signals: methods and applications},
  editor = {Stephan Baumann and Juan Jos\'{e} Burred and Andreas N\"{u}rnberger and Sebastian Stober},
  booktitle = {Proceedings of the 3rd Workshop on Learning the Semantics of Audio Signals (LSAS)},
  address = {Graz, Austria},
  month = {Dec},
  year = {2009},
  pages = {1},
  url = {http://lsas2009.dke-research.de/proceedings/lsas2009klapuri-abstract.pdf}
}
Abstract: We present a system that can learn effective classification models from music databases of very different characteristics, including both single-label collections indexed by genre or artist and multilabel databases of musical mood and instrumentation, where multiple tags can be applied to each track. Adaptability is attained by means of automatic feature and model selection, both embedded in the multiple-instance binary relevance learning of a Support Vector Machine. We discuss strategies for compensating overfitting and unbalanced training sets.
BibTeX:
@inproceedings{lsas2009burredPeeters,
  author = {Juan Jos\'{e} Burred and Geoffroy Peeters},
  title = {An Adaptive System for Music Classification and Tagging},
  editor = {Stephan Baumann and Juan Jos\'{e} Burred and Andreas N\"{u}rnberger and Sebastian Stober},
  booktitle = {Proceedings of the 3rd Workshop on Learning the Semantics of Audio Signals (LSAS)},
  address = {Graz, Austria},
  month = {Dec},
  year = {2009},
  pages = {3--16},
  url = {http://lsas2009.dke-research.de/proceedings/lsas2009burredPeeters.pdf}
}
Abstract: Thisstudyexploresreal-timevisualizationsoftimbre-related audio features using the music programming environment Max/MSP/Jitter. Following the approach of exploratory data-analysis, we present a palette of different visualization techniques, each one providing a different perspective on the feature-data. We introduce a simple notion of timbral distance, which can be used in real-time performance situations. The visualizations are further used to inform the control of audio effects by feature trajectories. Finally, we present a brief analysis of a section of Gerard Grisey's Modulations, which aims to combine score oriented methods with acoustical analysis of recordings.
BibTeX:
@inproceedings{lsas2009siedenburg,
  author = {Kai Siedenburg},
  title = {An Exploration of Real-Time Visualizations of Musical Timbre},
  editor = {Stephan Baumann and Juan Jos\'{e} Burred and Andreas N\"{u}rnberger and Sebastian Stober},
  booktitle = {Proceedings of the 3rd Workshop on Learning the Semantics of Audio Signals (LSAS)},
  address = {Graz, Austria},
  month = {Dec},
  year = {2009},
  pages = {17--30},
  url = {http://lsas2009.dke-research.de/proceedings/lsas2009siedenburg.pdf}
}
Abstract: Both low-level semantics of song texts and our emotional responses can be encoded in words. In order to model how we might perceive the emotional context of songs, we propose a simplified cognitive approach to bottom-up define term vector distances between lyrics and affective adjectives, which top-down constrain the latent semantics according to the psychological dimensions of valence and arousal. Projecting the lyrics and adjectives as vectors into a semantic space using LSA latent semantic analysis, their cosine similarities can be mapped as emotions over time. Subsequently we apply a three-way Tucker tensor decomposition to the derived LSA matrices, combined with a hierarchical Bayesian automatic relevance determination to find similarities across a selection of songs, and as a result identify two time series dramatic curvatures and three mixtures of affective components, which might function as emotional building blocks for generating the structure in lyrics.
BibTeX:
@inproceedings{lsas2009petersenMorupHansen,
  author = {Michael Kai Petersen and Morten M{\o}rup and Lars Kai Hansen},
  title = {Sparse but emotional decomposition of lyrics},
  editor = {Stephan Baumann and Juan Jos\'{e} Burred and Andreas N\"{u}rnberger and Sebastian Stober},
  booktitle = {Proceedings of the 3rd Workshop on Learning the Semantics of Audio Signals (LSAS)},
  address = {Graz, Austria},
  month = {Dec},
  year = {2009},
  pages = {31--43},
  url = {http://lsas2009.dke-research.de/proceedings/lsas2009petersenMorupHansen.pdf}
}
Abstract: In order to enrich music information retrieval applications with information about a user's listening habits, it is possible to automatically record a large variety of information about the listening context. However, recording such information may violate the user's privacy. This paper presents and discusses the results of a survey that has been conducted to assess the acceptance of listening context logging.
BibTeX:
@inproceedings{lsas2009stoberSteinbrecherNuernberger,
  author = {Sebastian Stober and Matthias Steinbrecher and Andreas N\"{u}rnberger},
  title = {A Survey on the Acceptance of Listening Context Logging for MIR Applications},
  editor = {Stephan Baumann and Juan Jos\'{e} Burred and Andreas N\"{u}rnberger and Sebastian Stober},
  booktitle = {Proceedings of the 3rd Workshop on Learning the Semantics of Audio Signals (LSAS)},
  address = {Graz, Austria},
  month = {Dec},
  year = {2009},
  pages = {45--57},
  url = {http://lsas2009.dke-research.de/proceedings/lsas2009stoberSteinbrecherNuernberger.pdf}
}
Abstract: This review article presents the state-of-the-art in context-based music similarity estimation. It gives an overview of different sources of context-based data on music entities and summarizes various approaches for constructing similarity measures based on the collaborative or cultural knowledge that is incorporated in these data sources. The strength of such context-based measures is elaborated as well as their drawbacks discussed.
BibTeX:
@inproceedings{lsas2009schedlKnees,
  author = {Markus Schedl and Peter Knees},
  title = {Context-based Music Similarity Estimation},
  editor = {Stephan Baumann and Juan Jos\'{e} Burred and Andreas N\"{u}rnberger and Sebastian Stober},
  booktitle = {Proceedings of the 3rd Workshop on Learning the Semantics of Audio Signals (LSAS)},
  address = {Graz, Austria},
  month = {Dec},
  year = {2009},
  pages = {59--74},
  url = {http://lsas2009.dke-research.de/proceedings/lsas2009schedlKnees.pdf}
}
Abstract: Considering that M.I.R. content-extraction algorithms are evaluated over annotated test-sets, it is worth discussing the robustness of the concepts used for these annotations. In this paper we discuss the robustness of local music annotations, more specifically "Music Structure" annotation. We define four conditions to be fulfilled by an annotation method to provide robust local annotation. We propose mathematical formulations of two of them. We then measure these criteria on existing "Music Structure" test-sets and discuss the pro's and con's of each test-set. From these, we derive a robust set of concepts which form a "multi-dimensional" description of the "Music Structure". We then apply this description to a set of 300 tracks representing various music genres and discuss the results.
BibTeX:
@inproceedings{lsas2009peetersDeruty,
  author = {Geoffroy Peeters and Emmanuel Deruty},
  title = {Is Music Structure Annotation Multi-Dimensional? A Proposal for Robust Local Music Annotation},
  editor = {Stephan Baumann and Juan Jos\'{e} Burred and Andreas N\"{u}rnberger and Sebastian Stober},
  booktitle = {Proceedings of the 3rd Workshop on Learning the Semantics of Audio Signals (LSAS)},
  address = {Graz, Austria},
  month = {Dec},
  year = {2009},
  pages = {75--90},
  url = {http://lsas2009.dke-research.de/proceedings/lsas2009peetersDeruty.pdf}
}
Abstract: In music information retrieval, acoustic low-level features are well studied and successfully applied in diverse classification tasks. So called mid-level features pose a very useful addition to low-level descriptors because they are specifically designed to bridge the gap between the low-level physical representation of music signals and the high-level semantic and symbolic information. Mid-level features have been proposed for different domains, such as dynamics, harmony or rhythm. Harmonic mid-level features, however, have mostly been used for harmonic analysis, such as key detection or chord boundary detection. They are rarely used as features for music classification. This publication describes different harmonic mid-level features and examines their usefulness for genre recognition. State of the art harmonic mid-level features are evaluated and their original extraction procedures are adapted to yield a satisfactory classification accuracy.
BibTeX:
@inproceedings{lsas2009gruhneDittmar,
  author = {Matthias Gruhne and Christian Dittmar},
  title = {Comparison of Harmonic Mid-level Representations for Genre Recognition},
  editor = {Stephan Baumann and Juan Jos\'{e} Burred and Andreas N\"{u}rnberger and Sebastian Stober},
  booktitle = {Proceedings of the 3rd Workshop on Learning the Semantics of Audio Signals (LSAS)},
  address = {Graz, Austria},
  month = {Dec},
  year = {2009},
  pages = {91--102},
  url = {http://lsas2009.dke-research.de/proceedings/lsas2009gruhneDittmar.pdf}
}