Beat this! Accurate beat tracking without DBN postprocessing
This addresses the need for more general and accurate beat tracking in music analysis, though it is incremental with some limitations in continuity and genre coverage.
The paper tackles the problem of beat and downbeat tracking in diverse music by removing the Dynamic Bayesian Network postprocessing and training on varied datasets, achieving a higher F1 score than the state of the art.
We propose a system for tracking beats and downbeats with two objectives: generality across a diverse music range, and high accuracy. We achieve generality by training on multiple datasets -- including solo instrument recordings, pieces with time signature changes, and classical music with high tempo variations -- and by removing the commonly used Dynamic Bayesian Network (DBN) postprocessing, which introduces constraints on the meter and tempo. For high accuracy, among other improvements, we develop a loss function tolerant to small time shifts of annotations, and an architecture alternating convolutions with transformers either over frequency or time. Our system surpasses the current state of the art in F1 score despite using no DBN. However, it can still fail, especially for difficult and underrepresented genres, and performs worse on continuity metrics, so we publish our model, code, and preprocessed datasets, and invite others to beat this.