Attention as a Perspective for Learning Tempo-invariant Audio Queries

TL;DR

This paper introduces a soft attention mechanism in audio--sheet music retrieval models to improve tempo-invariance, enabling more accurate retrieval across performances with varying tempos.

Contribution

It proposes a novel attention-based approach to address tempo variability in audio query retrieval, enhancing model robustness and performance.

Findings

Attention improves retrieval accuracy.

Model behavior aligns with musical intuition.

Empirical results show performance gains.

Abstract

Current models for audio--sheet music retrieval via multimodal embedding space learning use convolutional neural networks with a fixed-size window for the input audio. Depending on the tempo of a query performance, this window captures more or less musical content, while notehead density in the score is largely tempo-independent. In this work we address this disparity with a soft attention mechanism, which allows the model to encode only those parts of an audio excerpt that are most relevant with respect to efficient query codes. Empirical results on classical piano music indicate that attention is beneficial for retrieval performance, and exhibits intuitively appealing behavior.