Sebastian Böck, Florian Krebs, Gerhard Widmer,
"Joint Beat and Downbeat Tracking with Recurrent Neural Networks"
: Proceedings of the 17th International Society for Music Information Retrieval Conference (ISMIR), 8-2016
Joint Beat and Downbeat Tracking with Recurrent Neural Networks
Sprache des Titels:
Proceedings of the 17th International Society for Music Information Retrieval Conference (ISMIR)
JOINT BEAT AND DOWNBEAT TRACKING WITH RECURRENT
ock, Florian Krebs, and Gerhard Widmer
Department of Computational Perception
Johannes Kepler University Linz, Austria
In this paper we present a novel method for jointly extract-
ing beats and downbeats from audio signals. A recurrent
neural network operating directly on magnitude spectro-
grams is used to model the metrical structure of the audio
signals at multiple levels and provides an output feature
that clearly distinguishes between beats and downbeats.
A dynamic Bayesian network is then used to model bars
of variable length and align the predicted beat and down-
beat positions to the global best solution. We find that the
proposed model achieves state-of-the-art performance on a
wide range of different musical genres and styles.