Specific Music Transcription for Tutoring

An applicationspecific, musictranscription approach uses a customized human– computer interface to combine the strengths of humans and computers to enhance music transcription through instrument modeling and multimedia fusion. A utomatic music transcription (AMT) refers to the ability of computers to write note information—such as the pitch, onset time, duration, and source of each sound— after listening to the music. The ability to do AMT effectively is a Holy Grail for researchers working in this field. However, despite decades of research worldwide, a practically applicable, general-purpose transcription system doesn’t exist at present. It’s a problem that has been recognized by other researchers as well. Our application scenario is computer-assisted, musical-instrument tutoring, where the user has to practice most of the time without human coaching. The target users of our initial system are beginning violin students and singing students who are comfortable with computers. After consulting with a professional violin teacher and collaborators from our university’s Conservatory of Music, we tried the following three strategies to build a system that can be tested by users.