Tempo Estimation Based on Linear Prediction and Perceptual Modelling

Many applications demand the automatic induction of the tempo of a musical excerpt. The tempo estimation systems follow a general scheme that consists of two main steps: the creation of a feature list and the detection of periodicities on this list. In this study, we propose a new method for the implementation of the first step, along with the addition of a final step that will enhance the tempo estimation procedure. The proposed method for the extraction of the feature list is based on Gammatone subspace analysis and Linear Prediction Error Filters (LPEFs). As a final step on the system, the application of a model that approximates the tempo perception by human listeners is proposed. The results of the evaluation indicate the proposed method compares favourably with other, state-of-the-art tempo estimation methods, using only one frame of the musical experts when most of the literature methods demand the processing of the whole piece.