Ambisonics Super-Resolution Using A Waveform-Domain Neural Network

TL;DR

This paper introduces a neural network-based method that enhances first-order Ambisonics to higher orders, improving spatial accuracy and perceived audio quality without increasing channel count.

Contribution

A novel data-driven neural network approach that converts FOA to higher-order Ambisonics, surpassing traditional physics-based renderers in quality.

Findings

0.6dB reduction in positional mean squared error

80% improvement in perceived audio quality

Outperforms conventional renderers in quantitative and qualitative assessments

Abstract

Ambisonics is a spatial audio format describing a sound field. First-order Ambisonics (FOA) is a popular format comprising only four channels. This limited channel count comes at the expense of spatial accuracy. Ideally one would be able to take the efficiency of a FOA format without its limitations. We have devised a data-driven spatial audio solution that retains the efficiency of the FOA format but achieves quality that surpasses conventional renderers. Utilizing a fully convolutional time-domain audio neural network (Conv-TasNet), we created a solution that takes a FOA input and provides a higher order Ambisonics (HOA) output. This data driven approach is novel when compared to typical physics and psychoacoustic based renderers. Quantitative evaluations showed a 0.6dB average positional mean squared error difference between predicted and actual 3rd order HOA. The median qualitative rating showed an 80% improvement in perceived quality over the traditional rendering approach.