Self-Supervised Learning for Speech Enhancement through Synthesis

• Main Authors: Irvin, Bryce, Stamenovic, Marko, Kegler, Mikolaj, Yang, Li-Chia
• Format: Journal Article
• Language: English
• Published: 04.11.2022
• Subjects: Computer Science - Computation and Language; Computer Science - Learning; Computer Science - Sound

Modern speech enhancement (SE) networks typically implement noise suppression through time-frequency masking, latent representation masking, or discriminative signal prediction. In contrast, some recent works explore SE via generative speech synthesis, where the system's output is synthesized b...