# DFSNet: A Steerable Neural Beamformer Invariant to Microphone Array Configuration for Real-Time, Low-Latency Speech Enhancement

## TL;DR

DFSNet is a novel neural beamformer that remains invariant to microphone array configurations, enabling real-time, low-latency speech enhancement suitable for hearing aids by steering signals toward the source before beamforming.

## Contribution

This paper introduces DFSNet, a steerable neural beamformer invariant to microphone array geometry, simplifying reverberant speech enhancement in real-time applications.

## Key findings

- Achieves performance comparable to noncausal state-of-the-art methods.
- Operates with low latency, distortion, and computational load.
- Effective in reverberant and variable microphone configurations.

## Abstract

Invariance to microphone array configuration is a rare attribute in neural beamformers. Filter-and-sum (FS) methods in this class define the target signal with respect to a reference channel. However, this not only complicates formulation in reverberant conditions but also the network, which must have a mechanism to infer what the reference channel is. To address these issues, this study presents Delay Filter-and-Sum Network (DFSNet), a steerable neural beamformer invariant to microphone number and array geometry for causal speech enhancement. In DFSNet, acquired signals are first steered toward the speech source direction prior to the FS operation, which simplifies the task into the estimation of delay-and-summed reverberant clean speech. The proposed model is designed to incur low latency, distortion, and memory and computational burden, giving rise to high potential in hearing aid applications. Simulation results reveal comparable performance to noncausal state-of-the-art.