MetricGAN+: An Improved Version of MetricGAN for Speech Enhancement

Abstract

The discrepancy between the cost function used for training a speech enhancement model and human auditory perception usually makes the quality of enhanced speech unsatisfactory. Objective evaluation metrics which consider human perception can hence serve as a bridge to reduce the gap. Our previously proposed MetricGAN was designed to optimize objective metrics by connecting the metric with a discriminator. Because only the scores of the target evaluation functions are needed during training, the metrics can even be non-differentiable. In this study, we propose a MetricGAN+ in which three training techniques incorporating domain-knowledge of speech processing are proposed. With these techniques, experimental results on the VoiceBank-DEMAND dataset show that MetricGAN+ can increase PESQ score by 0.3 compared to the previous MetricGAN and achieve state-of-the-art results (PESQ score = 3.15).

Accepted by Interspeech 2021

References in corpus (1)

• MetricGAN: Generative Adversarial Networks based Black-box Metric Scores Optimization for Speech Enhancement

Cited by in corpus (5)

• SpeechBrain: A General-Purpose Speech Toolkit
• Glance and Gaze: A Collaborative Learning Framework for Single-channel Speech Enhancement
• Unsupervised Noise Adaptive Speech Enhancement by Discriminator-Constrained Optimal Transport
• TENET: A Time-reversal Enhancement Network for Noise-robust ASR
• Deep Learning-based Non-Intrusive Multi-Objective Speech Assessment Model with Cross-Domain Features