Music Tempo Estimation via Neural Networks -- A Comparative Analysis

TL;DR

This study compares neural network architectures for music tempo estimation, introducing a B-RNN model trained on a large dataset, and evaluates its performance against state-of-the-art models, highlighting the potential for improved accuracy especially on percussion tracks.

Contribution

It proposes and evaluates a B-RNN model for tempo estimation without auxiliary modules, providing a comprehensive comparison with CNN models on a large, diverse dataset.

Findings

B-RNN outperforms CNN on percussion-only tracks

Tempo estimation accuracy improves with percussion tracks

Large dataset enables detailed quantitative analysis

Abstract

This paper presents a comparative analysis on two artificial neural networks (with different architectures) for the task of tempo estimation. For this purpose, it also proposes the modeling, training and evaluation of a B-RNN (Bidirectional Recurrent Neural Network) model capable of estimating tempo in bpm (beats per minutes) of musical pieces, without using external auxiliary modules. An extensive database (12,550 pieces in total) was curated to conduct a quantitative and qualitative analysis over the experiment. Percussion-only tracks were also included in the dataset. The performance of the B-RNN is compared to that of state-of-the-art models. For further comparison, a state-of-the-art CNN was also retrained with the same datasets used for the B-RNN training. Evaluation results for each model and datasets are presented and discussed, as well as observations and ideas for future research. Tempo estimation was more accurate for the percussion only dataset, suggesting that the estimation can be more accurate for percussion-only tracks, although further experiments (with more of such datasets) should be made to gather stronger evidence.