Recent speech enhancement work, which makes use of neural networks trained with a loss derived in part using an adversarial metric prediction network, has shown to be very effective. However, by limiting the data used to train this metric prediction network to only the clean reference and the output of the speech enhancement network, only a limited range of the metric is learnt. Additionally, such speech enhancement systems are limited because they typically operate solely over magnitude spectrogram representations so they do not encode phase information. In this work, recent developments for phase-aware speech enhancement in such an adversarial framework are expanded in two ways to enable the metric prediction network to learn a full range of metric scores. Firstly, the metric predictor is also exposed to unenhanced ’noisy’ data during training. Furthermore, an additional network is introduced and trained alongside which attempts to produce outputs with a fixed ’lower’ target metric score, and expose the metric predictor to these ’de-enhanced’ outputs. It is found that performance increases versus a baseline system utilising a magnitude spectrogram speech enhancement network.

Authors: Close, George; Hain, Thomas; Goetze, Stefan
Affiliations: The University of Sheffield, UK; The University of Sheffield, UK; The University of Sheffield, UK(See document for exact affiliation information.)
AES Convention: 154 (May 2023) Paper Number: 10656
Publication Date: May 13, 2023 Import into BibTeX
Subject: Capturing
Permalink: https://www.aes.org/e-lib/browse.cfm?elib=22063

Download Now (522 KB)

This paper is Open Access which means you can download it for free.

Learn more about the AES E-Library

E-Library Location:   /conv/154/10656.pdf

Start a discussion about this Capturing!