Online Phase Reconstruction via DNN-based Phase Differences Estimation

Abstract

This paper presents a two-stage online phase reconstruction framework using causal deep neural networks (DNNs). Phase reconstruction is a task of recovering phase of the short-time Fourier transform (STFT) coefficients only from the corresponding magnitude. However, phase is sensitive to waveform shifts and not easy to estimate from the magnitude even with a DNN. To overcome this problem, we propose to use DNNs for estimating differences of phase between adjacent time-frequency bins. We show that convolutional neural networks are suitable for phase difference estimation, according to the theoretical relation between partial derivatives of STFT phase and magnitude. The estimated phase differences are used for reconstructing phase by solving a weighted least squares problem in a frame-by-frame manner. In contrast to existing DNN-based phase reconstruction methods, the proposed framework is causal and does not require any iterative procedure. The experiments showed that the proposed method outperforms existing online methods and a DNN-based method for phase reconstruction.

arXiv commentAccepted to IEEE/ACM Trans. Audio, Speech, and Language Processing

References in corpus (10)

• WaveNet: A Generative Model for Raw Audio
• Language Modeling with Gated Convolutional Networks
• On the Variance of the Adaptive Learning Rate and Beyond
• GANSynth: Adversarial Neural Audio Synthesis
• Phasebook and Friends: Leveraging Discrete Representations for Source Separation
• A Non-iterative Method for (Re)Construction of Phase from STFT Magnitude
• Phase recovery in NMF for audio source separation: an insightful benchmark
• A Deep Generative Model of Speech Complex Spectrograms
• Online Spectrogram Inversion for Low-Latency Audio Source Separation
• Generative adversarial network-based approach to signal reconstruction from magnitude spectrograms