DiffPhase: Generative Diffusion-based STFT Phase Retrieval

Abstract

Diffusion probabilistic models have been recently used in a variety of tasks, including speech enhancement and synthesis. As a generative approach, diffusion models have been shown to be especially suitable for imputation problems, where missing data is generated based on existing data. Phase retrieval is inherently an imputation problem, where phase information has to be generated based on the given magnitude. In this work we build upon previous work in the speech domain, adapting a speech enhancement diffusion model specifically for STFT phase retrieval. Evaluation using speech quality and intelligibility metrics shows the diffusion approach is well-suited to the phase retrieval task, with performance surpassing both classical and modern methods.

Accepted by ICASSP 2023

References in corpus (5)

• Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding
• Score-Based Generative Modeling through Stochastic Differential Equations
• Speech Enhancement and Dereverberation with Diffusion-based Generative Models
• A Study on Speech Enhancement Based on Diffusion Probabilistic Model
• Beyond Griffin-Lim: Improved Iterative Phase Retrieval for Speech

Cited by in corpus (2)

• DriftRec: Adapting diffusion models to blind JPEG restoration
• The Whole Is Greater than the Sum of Its Parts: Improving Music Source Separation by Bridging Network