Speech Enhancement Based On Cyclegan With Noise-informed Training

Abstract

Cycle-consistent generative adversarial networks (CycleGAN) were successfully applied to speech enhancement (SE) tasks with unpaired noisy-clean training data. The CycleGAN SE system adopted two generators and two discriminators trained with losses from noisy-to-clean and clean-to-noisy conversions. CycleGAN showed promising results for numerous SE tasks. Herein, we investigate a potential limitation of the clean-to-noisy conversion part and propose a novel noise-informed training (NIT) approach to improve the performance of the original CycleGAN SE system. The main idea of the NIT approach is to incorporate target domain information for clean-to-noisy conversion to facilitate a better training procedure. The experimental results confirmed that the proposed NIT approach improved the generalization capability of the original CycleGAN SE system with a notable margin.

Stats

• citations5
• S2 citations—
• github stars0
• HF likes0
• heat score5.84
• arxiv keyting2021speech

Related papers

• Conditional Generative Adversarial Networks For Speech Enhancement And Noise-robust Speaker Verification (2017)16.03
• Joint Magnitude Estimation And Phase Recovery Using Cycle-in-cycle GAN For Non-parallel Speech Enhancement (2021)7.81
• SEGAN: Speech Enhancement Generative Adversarial Network (2017)21.85
• A Multi-discriminator Cyclegan For Unsupervised Non-parallel Speech Domain Adaptation (2018)9.76
• Improving Noisy Student Training For Low-resource Languages In End-to-end ASR Using Cyclegan And Inter-domain Losses (2024)0.00
• Effective Noise-aware Data Simulation For Domain-adaptive Speech Enhancement Leveraging Dynamic Stochastic Perturbation (2024)3.58
• SEFGAN: Harvesting The Power Of Normalizing Flows And Gans For Efficient High-quality Speech Enhancement (2023)5.84
• Improving Semi-supervised End-to-end Automatic Speech Recognition Using Cyclegan And Inter-domain Losses (2022)3.58