Beat Tracking in Musical Recordings with Neural Networks
Loading...
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Beat tracking refers to the automated detection of the beat in musical recordings. Two neural network architectures are trained and evaluated for this task.
Description
The beat tracking process is divided into four stages: audio preprocessing, network training, peak-picking and beat detection. In the practical part, two network architectures (CNN and RNN) are trained and evaluated.