HGCN: Harmonic gated compensation network for speech enhancement

TL;DR

This paper introduces HGCN, a novel speech enhancement model that accurately detects and compensates for masked harmonics in noisy speech, outperforming existing methods.

Contribution

The paper proposes a harmonic gated compensation network with a high-resolution harmonic spectrum and gating mechanism for improved speech enhancement.

Findings

HGCN outperforms existing approaches in speech enhancement tasks.

The high-resolution harmonic spectrum improves harmonic location prediction.

The gating mechanism effectively refines enhancement results.

Abstract

Mask processing in the time-frequency (T-F) domain through the neural network has been one of the mainstreams for single-channel speech enhancement. However, it is hard for most models to handle the situation when harmonics are partially masked by noise. To tackle this challenge, we propose a harmonic gated compensation network (HGCN). We design a high-resolution harmonic integral spectrum to improve the accuracy of harmonic locations prediction. Then we add voice activity detection (VAD) and voiced region detection (VRD) to the convolutional recurrent network (CRN) to filter harmonic locations. Finally, the harmonic gating mechanism is used to guide the compensation model to adjust the coarse results from CRN to obtain the refinedly enhanced results. Our experiments show HGCN achieves substantial gain over a number of advanced approaches in the community.