An Information-theoretic Approach to Machine-oriented Music Summarization

TL;DR

This paper introduces an information-theoretic method for music summarization that minimizes relative entropy to improve machine-oriented processing, outperforming previous methods and addressing copyright concerns.

Contribution

It generalizes previous findings by evaluating music summarization through a probabilistic lens and proposes a new summarizer that minimizes relative entropy for better performance.

Findings

Relative entropy predicts summarization performance in bag-of-features tasks.

The proposed summarizer outperforms previous methods.

The approach reduces potential copyright issues.

Abstract

Music summarization allows for higher efficiency in processing, storage, and sharing of datasets. Machine-oriented approaches, being agnostic to human consumption, optimize these aspects even further. Such summaries have already been successfully validated in some MIR tasks. We now generalize previous conclusions by evaluating the impact of generic summarization of music from a probabilistic perspective. We estimate Gaussian distributions for original and summarized songs and compute their relative entropy, in order to measure information loss incurred by summarization. Our results suggest that relative entropy is a good predictor of summarization performance in the context of tasks relying on a bag-of-features model. Based on this observation, we further propose a straightforward yet expressive summarizer, which minimizes relative entropy with respect to the original song, that objectively outperforms previous methods and is better suited to avoid potential copyright issues.