Deep AHS: A Deep Learning Approach to Acoustic Howling Suppression

TL;DR

This paper introduces Deep AHS, a deep learning-based method that formulates acoustic howling suppression as a supervised speech separation task, enabling real-time, detection-free suppression with improved effectiveness.

Contribution

It proposes a novel deep learning approach using attention-based RNNs for acoustic howling suppression, converting it into a speech separation problem for better performance and training efficiency.

Findings

Effective suppression in various scenarios

Real-time streaming inference capability

Avoids explicit howling detection

Abstract

In this paper, we formulate acoustic howling suppression (AHS) as a supervised learning problem and propose a deep learning approach, called Deep AHS, to address it. Deep AHS is trained in a teacher forcing way which converts the recurrent howling suppression process into an instantaneous speech separation process to simplify the problem and accelerate the model training. The proposed method utilizes properly designed features and trains an attention based recurrent neural network (RNN) to extract the target signal from the microphone recording, thus attenuating the playback signal that may lead to howling. Different training strategies are investigated and a streaming inference method implemented in a recurrent mode used to evaluate the performance of the proposed method for real-time howling suppression. Deep AHS avoids howling detection and intrinsically prohibits howling from happening, allowing for more flexibility in the design of audio systems. Experimental results show the effectiveness of the proposed method for howling suppression under different scenarios.