Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Completing Sets of Prototype Transfer Functions for Subspace-based Direction of Arrival Estimation of Multiple Speakers

Daniel Fejgin, Simon Doclo

TL;DR

Experimental results for two speakers in noisy and reverberant environments clearly demonstrate that for all locations of the external microphone DOAs can be estimated more accurately with completed sets of prototype transfer functions than with incomplete sets.

Abstract

To estimate the direction of arrival (DOA) of multiple speakers, subspace-based prototype transfer function matching methods such as multiple signal classification (MUSIC) or relative transfer function (RTF) vector matching are commonly employed. In general, these methods require calibrated microphone arrays, which are characterized by a known array geometry or a set of known prototype transfer functions for several directions. In this paper, we consider a partially calibrated microphone array, composed of a calibrated binaural hearing aid and a (non-calibrated) external microphone at an unknown location with no available set of prototype transfer functions. We propose a procedure for completing sets of prototype transfer functions by exploiting the orthogonality of subspaces, allowing to apply matching-based DOA estimation methods with partially calibrated microphone arrays. For the MUSIC and RTF vector matching methods, experimental results for two speakers in noisy and reverberant environments clearly demonstrate that for all locations of the external microphone DOAs can be estimated more accurately with completed sets of prototype transfer functions than with incomplete sets. \c{opyright}20XX IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Completing Sets of Prototype Transfer Functions for Subspace-based Direction of Arrival Estimation of Multiple Speakers

TL;DR

Experimental results for two speakers in noisy and reverberant environments clearly demonstrate that for all locations of the external microphone DOAs can be estimated more accurately with completed sets of prototype transfer functions than with incomplete sets.

Abstract

To estimate the direction of arrival (DOA) of multiple speakers, subspace-based prototype transfer function matching methods such as multiple signal classification (MUSIC) or relative transfer function (RTF) vector matching are commonly employed. In general, these methods require calibrated microphone arrays, which are characterized by a known array geometry or a set of known prototype transfer functions for several directions. In this paper, we consider a partially calibrated microphone array, composed of a calibrated binaural hearing aid and a (non-calibrated) external microphone at an unknown location with no available set of prototype transfer functions. We propose a procedure for completing sets of prototype transfer functions by exploiting the orthogonality of subspaces, allowing to apply matching-based DOA estimation methods with partially calibrated microphone arrays. For the MUSIC and RTF vector matching methods, experimental results for two speakers in noisy and reverberant environments clearly demonstrate that for all locations of the external microphone DOAs can be estimated more accurately with completed sets of prototype transfer functions than with incomplete sets. \c{opyright}20XX IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
Paper Structure (10 sections, 20 equations, 2 figures)

This paper contains 10 sections, 20 equations, 2 figures.

Table of Contents

  1. Introduction
  2. Signal model and notation
  3. Subspace-based DOA Estimation Methods
  4. Construction of frequency-dependent SPS
  5. Fusion of frequency-dependent SPS
  6. Completing Sets of Prototype Transfer Functions
  7. Experimental results
  8. Experimental setup and implementation details
  9. Results
  10. Conclusions

Figures (2)

  • Figure 2: Average localization accuracy of DOA estimation using MUSIC (left) and the RTF vector matching method (right) with the conditions H/H (no eMic), H+E/H (eMic included only in the EVD of $\hat{\boldsymbol{\Phi}}_{\mathrm{y}}^{\mathrm{w}}$ but not in the prototype matching), and H+E/H+E (eMic included in the EVD of $\hat{\boldsymbol{\Phi}}_{\mathrm{y}}^{\mathrm{w}}$ and in the prototype matching).
  • Figure :