Deep Complex-Valued Neural Network-Based Triple-Path Mask and Steering Vector Estimation for Multichannel Target Speech Separation

We propose a deep complex-valued neural network-based beamforming framework for multichannel target speech separation. The deep complex-valued neural network predicts steering vectors and complex ratio masks for speaker signals. The masked signals are then used to calculate the spatial covariance ma...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of Signal Processing 2023/07/01, Vol.27(4), pp.87-91
Hauptverfasser: Qin, Mohan, Li, Li, Makino, Shoji
Format: Artikel
Sprache:eng ; jpn
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:We propose a deep complex-valued neural network-based beamforming framework for multichannel target speech separation. The deep complex-valued neural network predicts steering vectors and complex ratio masks for speaker signals. The masked signals are then used to calculate the spatial covariance matrices needed for minimum variance distortionless response (MVDR) beamforming. We propose triple-path modeling for mask estimation, which takes both intrachannel and interchannel features into consideration. Our experimental results revealed that the proposed framework achieves better target speech separation performance than do the baseline methods.
ISSN:1342-6230
1880-1013
DOI:10.2299/jsp.27.87