Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Real-time Speech Extraction Using Spatially Regularized Independent Low-rank Matrix Analysis and Rank-constrained Spatial Covariance Matrix Estimation

Yuto Ishikawa, Kohei Konaka, Tomohiko Nakamura, Norihiro Takamune, Hiroshi Saruwatari

TL;DR

This paper proposes the real-time extension of a speech extraction method based on independent low-rank matrix analysis (ILRMA) and rank-constrained spatial covariance matrix estimation (RCSCME).

Abstract

Real-time speech extraction is an important challenge with various applications such as speech recognition in a human-like avatar/robot. In this paper, we propose the real-time extension of a speech extraction method based on independent low-rank matrix analysis (ILRMA) and rank-constrained spatial covariance matrix estimation (RCSCME). The RCSCME-based method is a multichannel blind speech extraction method that demonstrates superior speech extraction performance in diffuse noise environments. To improve the performance, we introduce spatial regularization into the ILRMA part of the RCSCME-based speech extraction and design two regularizers. Speech extraction experiments demonstrated that the proposed methods can function in real time and the designed regularizers improve the speech extraction performance.

Real-time Speech Extraction Using Spatially Regularized Independent Low-rank Matrix Analysis and Rank-constrained Spatial Covariance Matrix Estimation

TL;DR

This paper proposes the real-time extension of a speech extraction method based on independent low-rank matrix analysis (ILRMA) and rank-constrained spatial covariance matrix estimation (RCSCME).

Abstract

Real-time speech extraction is an important challenge with various applications such as speech recognition in a human-like avatar/robot. In this paper, we propose the real-time extension of a speech extraction method based on independent low-rank matrix analysis (ILRMA) and rank-constrained spatial covariance matrix estimation (RCSCME). The RCSCME-based method is a multichannel blind speech extraction method that demonstrates superior speech extraction performance in diffuse noise environments. To improve the performance, we introduce spatial regularization into the ILRMA part of the RCSCME-based speech extraction and design two regularizers. Speech extraction experiments demonstrated that the proposed methods can function in real time and the designed regularizers improve the speech extraction performance.
Paper Structure (13 sections, 7 equations, 3 figures)

This paper contains 13 sections, 7 equations, 3 figures.

Table of Contents

  1. Introduction
  2. Related Offline methods
  3. ILRMA Kitamura2016TASLP
  4. RCSCME Kubo2020TASLP
  5. Proposed method
  6. Real-time RCSCME-based speech extraction method
  7. Incorporation of prior target speech direction information
  8. Spatial regularization using prior target steering vector
  9. ILRMA with null-based spatial regularization
  10. Experiments
  11. Experimental conditions
  12. Experimental results
  13. Conclusion

Figures (3)

  • Figure 1: Schematic of parallel processing in real-time RCSCME-based speech extraction method.
  • Figure 2: Boxplots of processing time for ILRMA part of all methods.
  • Figure 3: Boxplots of SDR improvement for all methods.