Text-Based Description of Music for Indexing, Retrieval, and Browsing
Supervisor(s) and Committee member(s): Dr. Gerhard Widmer (supervisor) Dr. Andreas Rauber (second reader)
The aim of this PhD thesis is to develop automatic methods that extract textual descriptions from the Web that can be associated with music pieces. Deriving descriptors for music permits to index large repositories with a diverse set of labels and allows for retrieving pieces and browsing collections. The techniques presented make use of common Web search engines to find related text content on the Web. From this content, descriptors are extracted that may serve as
- labels that facilitate orientation within browsing interfaces to music collections, especially in a three-dimensional browsing interface presented,
- indexing terms, used as features in music retrieval systems that can be queried using descriptive free-form text as input, and
- features in adaptive retrieval systems that aim at providing more user-targeted recommendations based on the user’s searching behaviour for exploration of music collections.
In the context of this thesis, different extraction, indexing, and retrieval strategies are elaborated and evaluated. Furthermore, the potential of complementing Web-based retrieval with acoustic similarity extracted from the audio signal, as well as complementing audio-similarity-based browsing approaches with Web-based descriptors is investigated and demonstrated in prototype applications.
Department of Computational Perception, Johannes Kepler University Linz, Austria
The Department of Computational Perception of the Johannes Kepler University Linz, Austria carries out basic and applied research in machine learning, pattern recognition, knowledge extraction, information retrieval, and generally Artificial and Computational Intelligence with a focus on intelligent audio (specifically: music) and image processing. Headed by Prof. Gerhard Widmer, it has become one of the world-leading research groups in Music Information Retrieval. Current music-related research directions comprise the recognition and transcription of musical dimensions such as beat, tempo, and pitch from audio recordings, real-time tracking of scores and vocals from live performances, automatic rendering of expressive piano performances, music retrieval and recommendation in collections of millions of songs, and the development of novel interfaces to music collections. In addition to signal-based music research, a focus is also put on Web-mining techniques to exploit contextually related information on music.
The Department of Computational Perception maintains close cooperation links with the Austrian Research Institute for Artificial Intelligence (OFAI), Vienna, and in particular with its Machine Learning, Data Mining, and Intelligent Music Processing Group (which is also headed by Prof. Gerhard Widmer).