Recent research in computational groove estimation and beat tracking from audio
A common response when experiencing music is to move in sync to it, e.g. tap your foot or nod your head in time to the beat, or dance. This phenomenon raises several questions of special interest, e.g. which musical properties can explain physical entrainment to the music? Which are their physical correlates in the sound signal? How can synchronization to the musical beat be modeled computationally? And how can the increasing availability of data and algorithms help leverage these issues? In this talk I will propose a survey of some recent research conducted on computational groove estimation and beat tracking from sound signals. I will also put a special focus on the important issue of algorithm evaluation in the absence of reference annotations.