Tempo-Invariant Processing of Rhythm with Convolutional Neural Networks
This work addresses a key problem in music information retrieval for systems that need to handle tempo variations, though it appears incremental as it builds on existing tempo-invariant methods.
The paper tackles the challenge of processing rhythm patterns across varying tempi in music information retrieval by proposing a log-frequency representation of rhythm-related activations for convolutional neural networks, which promotes tempo invariance and can be applied to tasks like tempo estimation and beat tracking.
Rhythm patterns can be performed with a wide variation of tempi. This presents a challenge for many music information retrieval (MIR) systems; ideally, perceptually similar rhythms should be represented and processed similarly, regardless of the specific tempo at which they were performed. Several recent systems for tempo estimation, beat tracking, and downbeat tracking have therefore sought to process rhythm in a tempo-invariant way, often by sampling input vectors according to a precomputed pulse level. This paper describes how a log-frequency representation of rhythm-related activations instead can promote tempo invariance when processed with convolutional neural networks. The strategy incorporates invariance at a fundamental level and can be useful for most tasks related to rhythm processing. Different methods are described, relying on magnitude, phase relationships of different rhythm channels, as well as raw phase information. Several variations are explored to provide direction for future implementations.