Blind Estimation of Sub-band Acoustic Parameters from Ambisonics Recordings using Spectro-Spatial Covariance Features

TL;DR

This paper presents a novel spectro-spatial covariance feature and a deep learning framework for blind estimation of acoustic parameters from Ambisonics recordings, significantly improving accuracy over existing methods.

Contribution

It introduces the Spectro-Spatial Covariance Vector (SSCV) feature and FOA-Conv3D network, advancing blind acoustic parameter estimation from Ambisonics data.

Findings

Over 50% reduction in estimation errors for T60, DRR, and C50.

SSCV feature outperforms spectral-only features.

FOA-Conv3D achieves higher variance explained than CNN and CRNN.

Abstract

Estimating frequency-varying acoustic parameters is essential for enhancing immersive perception in realistic spatial audio creation. In this paper, we propose a unified framework that blindly estimates reverberation time (T60), direct-to-reverberant ratio (DRR), and clarity (C50) across 10 frequency bands using first-order Ambisonics (FOA) speech recordings as inputs. The proposed framework utilizes a novel feature named Spectro-Spatial Covariance Vector (SSCV), efficiently representing temporal, spectral as well as spatial information of the FOA signal. Our models significantly outperform existing single-channel methods with only spectral information, reducing estimation errors by more than half for all three acoustic parameters. Additionally, we introduce FOA-Conv3D, a novel back-end network for effectively utilising the SSCV feature with a 3D convolutional encoder. FOA-Conv3D outperforms the convolutional neural network (CNN) and recurrent convolutional neural network (CRNN) backends, achieving lower estimation errors and accounting for a higher proportion of variance (PoV) for all 3 acoustic parameters.