BAE-Net: A Low complexity and high fidelity Bandwidth-Adaptive neural network for speech super-resolution

TL;DR

BAE-Net is a novel neural network designed for speech bandwidth extension that adaptively handles varying bandwidths, improving speech quality and efficiency, especially suitable for edge devices.

Contribution

This paper introduces BAE-Net, a bandwidth-adaptive neural network for speech super-resolution that addresses fluctuating bandwidths and includes a lightweight version for edge applications.

Findings

BAE-Net outperforms existing methods in speech quality and efficiency.

The dual-stream architecture effectively reconstructs high-frequency speech content.

BAE-Net-lite offers a lightweight solution suitable for real-time edge deployment.

Abstract

Speech bandwidth extension (BWE) has demonstrated promising performance in enhancing the perceptual speech quality in real communication systems. Most existing BWE researches primarily focus on fixed upsampling ratios, disregarding the fact that the effective bandwidth of captured audio may fluctuate frequently due to various capturing devices and transmission conditions. In this paper, we propose a novel streaming adaptive bandwidth extension solution dubbed BAE-Net, which is suitable to handle the low-resolution speech with unknown and varying effective bandwidth. To address the challenges of recovering both the high-frequency magnitude and phase speech content blindly, we devise a dual-stream architecture that incorporates the magnitude inpainting and phase refinement. For potential applications on edge devices, this paper also introduces BAE-NET-lite, which is a lightweight, streaming and efficient framework. Quantitative results demonstrate the superiority of BAE-Net in terms of both performance and computational efficiency when compared with existing state-of-the-art BWE methods.