BEHM-GAN: Bandwidth Extension of Historical Music using Generative Adversarial Networks

TL;DR

BEHM-GAN is a novel generative adversarial network model designed to extend the audio bandwidth of historical music recordings, significantly improving perceptual sound quality and aiding in data-driven music restoration.

Contribution

This paper introduces BEHM-GAN, a GAN-based approach that effectively extends bandwidth in real historical recordings using spectrograms and regularization, outperforming existing methods.

Findings

Outperforms baseline methods in objective and subjective tests.

Significantly improves perceptual sound quality in early-20th-century recordings.

Substantial increase in mean opinion scores after bandwidth extension.

Abstract

Audio bandwidth extension aims to expand the spectrum of narrow-band audio signals. Although this topic has been broadly studied during recent years, the particular problem of extending the bandwidth of historical music recordings remains an open challenge. This paper proposes BEHM-GAN, a model based on generative adversarial networks, as a practical solution to this problem. The proposed method works with the complex spectrogram representation of audio and, thanks to a dedicated regularization strategy, can effectively extend the bandwidth of out-of-distribution real historical recordings. The BEHM-GAN is designed to be applied as a second step after denoising the recording to suppress any additive disturbances, such as clicks and background noise. We train and evaluate the method using solo piano classical music. The proposed method outperforms the compared baselines in both objective and subjective experiments. The results of a formal blind listening test show that BEHM-GAN significantly increases the perceptual sound quality in early-20th-century gramophone recordings. For several items, there is a substantial improvement in the mean opinion score after enhancing historical recordings with the proposed bandwidth-extension algorithm. This study represents a relevant step toward data-driven music restoration in real-world scenarios.