Automatic Music Transcription


Convert music audio signals from smartphones and tablets into sheet music for each of the instruments involved.


Allow musicians to record live music performances in noisy environments such as coffee houses and create music transcripts such as sheet music, guitar tabs, or MIDI files, for each of the instruments. Create software allowing musicians to document their own original performances and to search for information on cover songs when the performed version differs significantly from the standard recorded version.

Key Elements:

Digital signal processing, personal mobile device programming, music recording and production technology, music composition and transcription, physics of musical instruments, psychology of music perception, acoustics.

Research Issues:

Instrument signal separation, crowd noise reduction, out-of-tune instruments, processing power limitations of personal mobile devices, music information display on small touchscreens, modelling of human music transcription for application to automatic transcription, embedding knowledge of music composition in automatic transcription systems.

Project Partners:


Desired Majors, Preparations, Interests:

Music – composition, transcription, recording, production, MIDI, electronic music
Electrical Engineering – digital signal processing, image processing
Computer Science – personal mobile device programming, cloud computing, data visualization
Physics – acoustics, physical modelling of musical instruments



Contact: Jennifer Smith –