Search papers, labs, and topics across Lattice.
This paper addresses the issue of invalid outputs in neural networks for beat tracking, which often produce consecutive downbeats and erratic tempo changes. By hypothesizing that these inconsistencies arise from inadequate modeling of multiple plausible beat grids, the authors introduce a masked diffusion approach that iteratively refines predictions. The proposed method significantly reduces erratic behaviors and enhances overall beat-tracking performance through innovative modifications to standard masked diffusion techniques.
Erratic beat tracking outputs can be effectively mitigated by a novel masked diffusion approach that models multiple plausible interpretations.
Current neural networks for beat tracking generate invalid outputs, such as consecutive downbeats and erratic tempo changes, even when these are not present in the training data. Heavy post-processing techniques can alleviate these problems, but the original cause of this inconsistent behaviour remains unknown. We hypothesise that it stems from inadequate modelling of multiple plausible output beat grids, resulting in an invalid mixture of competing interpretations. We propose a masked diffusion approach that properly models multiple outputs and enables the model to build coherent predictions through iterative inference. We devise three modifications to standard masked diffusion that enable its application to beat tracking: independent masking of beats and downbeats during training and inference, a balanced masking scheduler for inference, and peak-picking across inference steps. Our approach reduces erratic behaviours and improves beat-tracking performance.
