Getting a computer to "feel the beat" of live music, so that it can clap along or dance or play accompaniment, is devilishly difficult... but why? We'll examine a beat-tracking system which performs comparatively well (but still miserably) for a limited range of pop music inputs — and contemplate what it takes to do better.
Presenter: Marvin Humphrey
Paper: "An Audio-based Real-time Beat Tracking System for Music With or Without Drum-sounds", by Masataka Goto (2001)
(Reading the paper is not required — it is expected that many attendees will have at most glanced at the abstract.)
Section 5, "Applications", provides motivation — why solving the beat-tracking problem well is worthwhile. This presentation will focus on adaptive accompaniment: what it takes to have a computer play along in time to a live performance.
Section 2, "Beat-tracking problem (inverse problem)", provides a succinct and insightful problem definition.
Much of the rest of the paper focuses on engineering challenges, such as extracting kick and snare drum signals and identifying when chord changes occur. We will explore how these techniques might be augmented or adapted to achieve near 100% reliability.
Street parking on 6th, 7th & 8th Avenues north of B Street is usually easy at that hour. Meters nearby are free after 6. Read signage before you park on A street.
If you're interested in presenting a paper please fill out this form (https://docs.google.com/forms/d/e/1FAIpQLScaI-fWdys27-ByT_HdtsJ73V4AxZr0hf1GSqLsQ1IwAaPdIQ/viewform) or talk to us in person at the meetup.