Back to publications
Enhancement

GSE-Net: Multi-Scale Feature Extraction for Single-Channel Speech Enhancement

Authors: Alireza Kayed-Nezami, Rahil Mahdian Toroghi, Hassan Zareian

2026 34th International Conference on Electrical Engineering (ICEE) · 2026

Abstract

This paper introduces GSE-Net, a deep hybrid architecture that combines the local feature modeling capabilities of convolutional networks with the long-range dependency modeling of attention-based modules. This design strengthens the model's capacity to jointly capture global and local structures. Experimental evaluations on the Voice-Bank+DEMAND dataset show that GSE-Net delivers results that are comparable to, and in many cases surpass, leading state-of-the-art methods.