Employing Real Training Data for Deep Noise Suppression

TL;DR

This paper introduces PESQ-DNN, a neural network that estimates speech quality scores without needing clean speech references, enabling the use of real-world data for training deep noise suppression models and improving their performance.

Contribution

The paper presents a novel reference-free perceptual loss using PESQ-DNN and an epoch-wise training protocol to effectively incorporate real data into DNS model training.

Findings

Outperforms synthetic-only training methods in PESQ scores.

Significantly surpasses Interspeech 2021 DNS Challenge baseline.

Achieves higher DNSMOS scores on real and synthetic data.

Abstract

Most deep noise suppression (DNS) models are trained with reference-based losses requiring access to clean speech. However, sometimes an additive microphone model is insufficient for real-world applications. Accordingly, ways to use real training data in supervised learning for DNS models promise to reduce a potential training/inference mismatch. Employing real data for DNS training requires either generative approaches or a reference-free loss without access to the corresponding clean speech. In this work, we propose to employ an end-to-end non-intrusive deep neural network (DNN), named PESQ-DNN, to estimate perceptual evaluation of speech quality (PESQ) scores of enhanced real data. It provides a reference-free perceptual loss for employing real data during DNS training, maximizing the PESQ scores. Furthermore, we use an epoch-wise alternating training protocol, updating the DNS model on real data, followed by PESQ-DNN updating on synthetic data. The DNS model trained with the PESQ-DNN employing real data outperforms all reference methods employing only synthetic training data. On synthetic test data, our proposed method excels the Interspeech 2021 DNS Challenge baseline by a significant 0.32 PESQ points. Both on synthetic and real test data, the proposed method beats the baseline by 0.05 DNSMOS points - although PESQ-DNN optimizes for a different perceptual metric.