Extended Vector-Based EB-ESPRIT Method

Byeongho Jo,Jung-Woo Choi,Franz Zotter

doi:10.1109/taslp.2020.2996090

Abstract

The estimation of direction of arrivals (DoAs) from spherical microphone array data is one of the key issues in extracting source information from all-around audio recordings. One such technique is the eigenbeam estimation of signal parameters via the rotational invariance technique (EB-ESPRIT), which separates the signal subspace related to the stationary sound field and then directly estimates DoAs of multiple sound sources. EB-ESPRIT has been evolved in many different ways by involving different types of recurrence relations of spherical harmonics, all of which are able to identify DoAs of a limited number of sources that are noticeably smaller than the number of finite-order spherical harmonic coefficients recorded. In this work, we report that it is possible to go beyond the known limits of detectable sources. The proposed formula is also based on conventional recurrence relations and probably permits to reach the ultimate limit by additional constraints of the signal parameters that can better exploit the highest-order coefficients. Monte-Carlo simulations conducted with various source positions and signal-to-noise ratios (SNRs) reveal that the proposed technique can detect more sources with insignificant loss in estimation performance and robustness.

Highlights

E STIMATION of direction of arrivals (DoAs) is an important topic in sound source localization (SSL) problems
In the case with aliasing (Fig. 3(b)), the performance ranking is the same for the low signal-to-noise ratios (SNRs) conditions (SNR = 0, 5 dB), but it can be seen that the root-mean-square error (RMSE) of the proposed method are slightly higher than those of the EB-ESPRIT and vector-based EB-ESPRIT for very high SNRs greater than 10 dB
It can be seen that the proposed method can estimate up to N 2 + 4N/3 with negligible amount of errors. These results show that the number of detectable sources using the proposed EB-ESPRIT doubles for the first-order Ambisonics signal (N = 1), and it increases from 9 to 13 in the case of third-order Ambisonics (N = 3), which is currently popular in studies using spherical arrays

Summary

Introduction

E STIMATION of direction of arrivals (DoAs) is an important topic in sound source localization (SSL) problems. DoAs of noise sources are important information in noise control problems. In the speech recognition task, locations of speakers are essential prerequisites for the source separation and noise suppression [1], [2]. For spatial audio coding and parameterization, DoAs are primary cues for separating the directional audio component from ambient signals, which enable the compression of multichannel data or optimal mixing to playback signals for various loudspeaker layouts [3]. There have been attempts to identify wall locations using echoes, and the DoA estimation can provide directions of image sources producing echoes inside a room [4]–[6]. Techniques for the DoA estimation can be largely categorized as parametric and non-parametric techniques [7]. The non-parametric technique generates a map of steered response power or detection probability, and peaks of high power are JO et al.: EXTENDED VECTOR-BASED EB-ESPRIT METHOD

Methods

Results

Conclusion