EBEN: Extreme bandwidth extension network applied to speech signals captured with noise-resilient body-conduction microphones

TL;DR

This paper introduces EBEN, a GAN-based model that enhances noisy, limited-bandwidth speech captured with body-conduction microphones, achieving real-time, state-of-the-art wideband speech recovery.

Contribution

The paper proposes a novel multiband decomposition approach combined with a U-Net-like GAN architecture for effective speech bandwidth extension from body-conduction microphone signals.

Findings

Achieves state-of-the-art speech enhancement results.

Operates in real-time with a lightweight model.

Effectively recovers wideband speech from noise-resilient recordings.

Abstract

In this paper, we present Extreme Bandwidth Extension Network (EBEN), a Generative Adversarial network (GAN) that enhances audio measured with body-conduction microphones. This type of capture equipment suppresses ambient noise at the expense of speech bandwidth, thereby requiring signal enhancement techniques to recover the wideband speech signal. EBEN leverages a multiband decomposition of the raw captured speech to decrease the data time-domain dimensions, and give better control over the full-band signal. This multiband representation is fed to a U-Net-like model, which adopts a combination of feature and adversarial losses to recover an enhanced audio signal. We also benefit from this original representation in the proposed discriminator architecture. Our approach can achieve state-of-the-art results with a lightweight generator and real-time compatible operation.