Mask-dependent Phase Estimation for Monaural Speaker Separation

TL;DR

This paper introduces a phase estimation network that improves monaural speaker separation by predicting phase information based on a T-F mask, addressing phase mismatch issues in traditional methods.

Contribution

It proposes a mask-dependent permutation invariant training criterion and an inverse mask weighted loss for effective phase prediction in speaker separation.

Findings

Achieves comparable performance to iterative phase reconstruction methods.

Simplifies phase estimation process in monaural speaker separation.

Demonstrates effectiveness on WSJ0-2mix dataset.

Abstract

Speaker separation refers to isolating speech of interest in a multi-talker environment. Most methods apply real-valued Time-Frequency (T-F) masks to the mixture Short-Time Fourier Transform (STFT) to reconstruct the clean speech. Hence there is an unavoidable mismatch between the phase of the reconstruction and the original phase of the clean speech. In this paper, we propose a simple yet effective phase estimation network that predicts the phase of the clean speech based on a T-F mask predicted by a chimera++ network. To overcome the label-permutation problem for both the T-F mask and the phase, we propose a mask-dependent permutation invariant training (PIT) criterion to select the phase signal based on the loss from the T-F mask prediction. We also propose an Inverse Mask Weighted Loss Function for phase prediction to focus the model on the T-F regions in which the phase is more difficult to predict. Results on the WSJ0-2mix dataset show that the phase estimation network achieves comparable performance to models that use iterative phase reconstruction or end-to-end time-domain loss functions, but in a more straightforward manner.