Parallel Gated Neural Network With Attention Mechanism For Speech Enhancement

TL;DR

This paper introduces a novel monaural speech enhancement system using a sequence-to-sequence architecture with gated neural networks and attention, improving speech quality by capturing long-term contextual information.

Contribution

It proposes a new speech enhancement model combining Feature Extraction, Compensation, and Mask Blocks with attention mechanisms, outperforming recent models on standard datasets.

Findings

Achieved higher ESTOI scores than recent models.

Obtained better PESQ scores indicating improved speech quality.

Demonstrated effectiveness on the Librispeech dataset.

Abstract

Deep learning algorithm are increasingly used for speech enhancement (SE). In supervised methods, global and local information is required for accurate spectral mapping. A key restriction is often poor capture of key contextual information. To leverage long-term for target speakers and compensate distortions of cleaned speech, this paper adopts a sequence-to-sequence (S2S) mapping structure and proposes a novel monaural speech enhancement system, consisting of a Feature Extraction Block (FEB), a Compensation Enhancement Block (ComEB) and a Mask Block (MB). In the FEB a U-net block is used to extract abstract features using complex-valued spectra with one path to suppress the background noise in the magnitude domain using masking methods and the MB takes magnitude features from the FEBand compensates the lost complex-domain features produced from ComEB to restore the final cleaned speech. Experiments are conducted on the Librispeech dataset and results show that the proposed model obtains better performance than recent models in terms of ESTOI and PESQ scores.