DFingerNet: Noise-Adaptive Speech Enhancement for Hearing Aids

TL;DR

DFingerNet enhances hearing aid speech processing by integrating noise-adaptive techniques into DeepFilterNet, leveraging background recordings for improved denoising performance with minimal extra computation.

Contribution

This paper introduces DFingerNet, a noise-adaptive extension of DeepFilterNet that utilizes background recordings for in-context adaptation, improving speech enhancement in hearing aids.

Findings

Superior performance on DNS Challenge benchmarks

Effective noise adaptation with minimal computational overhead

Improved generalisability across diverse environments

Abstract

The DeepFilterNet (DFN) architecture was recently proposed as a deep learning model suited for hearing aid devices. Despite its competitive performance on numerous benchmarks, it still follows a `one-size-fits-all' approach, which aims to train a single, monolithic architecture that generalises across different noises and environments. However, its limited size and computation budget can hamper its generalisability. Recent work has shown that in-context adaptation can improve performance by conditioning the denoising process on additional information extracted from background recordings to mitigate this. These recordings can be offloaded outside the hearing aid, thus improving performance while adding minimal computational overhead. We introduce these principles to the DFN model, thus proposing the DFingerNet (DFiN) model, which shows superior performance on various benchmarks inspired by the DNS Challenge.