The topic of drum transcription -- of identifying what drums are playing when in an audio signal -- has received considerable attention from researchers in music information retrieval. Transcription models incorporating an awareness of position in the bar or metric hierarchy are less common, however, despite findings in music psychology to suggest that such an awareness is a useful reality. This paper presents a system for drum transcription in polyphonic music, which is shown to be comparable with state-of-the-art approaches. Importantly, the system also conducts a synchronous transcription of bar position, meaning that, unlike other drum transcription systems, its output can be used directly for subsequent analyses leveraging positional information (e.g., the discovery of repeated rhythmic patterns).
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.