Cano, E.E.CanoDittmar, C.C.DittmarAbeßer, J.J.AbeßerKehling, C.C.KehlingGrollmisch, S.S.Grollmisch2022-03-062024-04-152022-03-062018https://publica.fraunhofer.de/handle/publica/26155910.1007/978-3-662-55004-5_41In this chapter, the application of music information retrieval (MIR ) technologies in the development of music education tools is addressed. First, the relationship between technology and music education is described from a historical point of view, starting with the earliest attempts to use audio technology for education and ending with the latest developments and current research conducted in the field. Second, three MIR technologies used within a music education context are presented: 1. The use of pitch-informed solo and accompaniment separation as a tool for the creation of practice content 2. Drum transcription for real-time music practice 3. Guitar transcription with plucking style and expression style detection. In each case, proposed methods are clearly described and evaluated. Objective perceptual quality metrics were used to evaluate the proposed method for solo/accompaniment separation. Mean overall perceptual scores (OPS ) of 24.68 and 34.68 were obtained for the solo and accompaniment tracks respectively. These scores are on par with the state-of-the-art methods with respect to perceptual quality of separated music signals. A dataset of 17 real-world multitrack recordings was used for evaluation. In the drum sound detection task, an F-measure of 0.96 was obtained for snare drum, kick drum, and hi-hat detection. For this evaluation, a dataset of 30 manually annotated real-world drum loops with an onset tolerance of 50 ms was used. For the guitar plucking style and guitar expression style detection tasks, F-measures of 0.93 and 0.83 were obtained respectively. For this evaluation, a dataset containing 261 recordings of both isolated notes as well as monophonic and polyphonic melodies with note-wise annotations was used. To conclude the chapter, the remaining challenges that need to be addressed to more effectively use MIR technologies in the development of music education applications are described.enAutomatic Music Analysis621006book article