Deep Learning Based Phase Reconstruction for Speaker Separation: A Trigonometric Perspective

By Zhong-Qiu Wang, Ke Tan, Deliang Wang. 2018

100 citations recorded by the source corpus.

This study investigates phase reconstruction for deep learning based monaural talker-independent speaker separation in the short-time Fourier transform (STFT) domain. The key observation is that, for a mixture oftwo sources, with their magnitudes accurately estimated and under a geometric constraint, the absolute phase difference between each source and the mixture can be uniquely determined; in addition, the source phases at each time-frequency $({\text{T}} - {{\Gamma }})$ unit can be narrowed down to only two candidates. To pick the right candidate, we propose three algorithms based on iterative phase reconstruction, group delay estimation, and phase-difference sign prediction. State-of-the-art results are obtained on the publicly available wsj0-2mix and 3 mix corpus.

Subfields: Artificial Intelligence, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Signal Processing

Citations in the corpus

Single Channel Voice Separation for Unknown Number of Speakers Under Reverberant and Noisy Settings (2020, methodology): “In this work, we focus on fully supervised voice separation using a single microphone, which has seen a great leap in performance following the recent success of deep learning models considering both frequency domain [1, 2, 7, 8, 9, 10], and time-domain [11, 12, 13, 14, 15, 16] modeling.”
Voice Separation with an Unknown Number of Multiple Speakers (2020, methodology): “Recently, Conv-Tasnet was scaled by Zhang et al. (2020), who proposed to train several separator networks in parallel to perform an ensemble.”
Dual-Path RNN: Efficient Long Sequence Modeling for Time-Domain Single-Channel Speech Separation (2019, neutral): “l size SI-SNRi (dB) SDRi (dB) DPCL++ [31] 13.6M 10.8 – uPIT-BLSTM-ST [29] 92.7M – 10.0 ADANet [32] 9.1M 10.4 10.8 WA-MISI-5 [33] 32.9M 12.6 13.1 Conv-TasNet-gLN [4] 5.1M 15.3 15.6 Sign Prediction Net [34] 55.2M 15.3 15.6 Deep CASA [20] 12.8M 17.7 18.0 FurcaNeXt [5] 51.4M – 18.4 DPRNN-TasNet 2.6M 18.”
WHAM!: Extending Speech Separation to Noisy Environments (2019, background): “This has brought forth a novel class of network architectures and objective functions beneﬁting from some type of phase processing, either implicitly by directly optimizing the time domain waveform [9, 13, 14], or explicitly via phase estimation algorithms [8, 13, 15].”
Phase-aware Single-stage Speech Denoising and Dereverberation with U-Net (2020, methodology): “Inspired by this, [18] proposed to estimate a rotational direction of the absolute phase difference using a sign-prediction network.”
Attention Is All You Need In Speech Separation (2020, unknown): “io (SI-SNR) [27] with utterance-level permutation invariant loss [13]. Table 1. Best results on the WSJ0-2mix dataset. Model SI-SNRi SDRi # Param Stride Tasnet [29] 10.8 11.1 n.a 20 SignPredictionNet [30] 15.3 15.6 55.2M 8 ConvTasnet [15] 15.3 15.6 5.1M 10 Two-Step CTN [31] 16.1 n.a. 8.6M 10 DeepCAS”
Divide and Conquer: A Deep CASA Approach to Talker-Independent Monaural Speaker Separation (2019, background): “In [41], a phase prediction network is further added to [40] to estimate the clean phase of each speaker source.”