Using Statistic Model to Capture the Association between Timbre and Perceived Tempo

The estimation of the perceived tempo is required in many MIR applications. However, automatic tempo estimation itself is still an open problem due to the insufficient understanding of the inherent mechanisms of the tempo perception. Published methods only use the information of rhythm pattern, so they may meet the half/double tempo error problem. To solve this problem, We propose to use statistic model to investigate the association between timbre and tempo and use timbre information to improve the performance of tempo estimation. Experiment results show that this approach performs at least comparably to existing tempo extraction algorithms.