2. PARAMETERS

In this paper, we present a Hidden Markov Model (HMM) based beat tracking system that simultaneously extracts downbeats, beat times, tempo, meter and rhythmic pat-terns. Our model builds upon the basic structure proposed by Whiteley et. al [7], which we further modified by in-troducing a new observation model: rhythmic patterns are learned directly from data, which makes the model adapt-able to the rhythmical structure of any kind of music. The MIREX beat tracking evaluation- 30 results using ten mea-sures and three datasets- placed our algorithm among the top ten performing algorithms 18 times