TS-URGENet: A Three-stage Universal Robust and Generalizable Speech Enhancement Network

TL;DR

TS-URGENet is a novel three-stage speech enhancement network designed to handle various distortions and input formats, improving speech quality and robustness across diverse conditions.

Contribution

The paper introduces a three-stage architecture for universal speech enhancement, addressing multiple distortions simultaneously, which is a new approach in the field.

Findings

Ranked 2nd in Interspeech 2025 URGENT Challenge Track 1.

Effectively mitigates packet loss, noise, reverberation, and bandwidth limitations.

Demonstrates superior performance over existing methods.

Abstract

Universal speech enhancement aims to handle input speech with different distortions and input formats. To tackle this challenge, we present TS-URGENet, a Three-Stage Universal, Robust, and Generalizable speech Enhancement Network. To address various distortions, the proposed system employs a novel three-stage architecture consisting of a filling stage, a separation stage, and a restoration stage. The filling stage mitigates packet loss by preliminarily filling lost regions under noise interference, ensuring signal continuity. The separation stage suppresses noise, reverberation, and clipping distortion to improve speech clarity. Finally, the restoration stage compensates for bandwidth limitation, codec artifacts, and residual packet loss distortion, refining the overall speech quality. Our proposed TS-URGENet achieved outstanding performance in the Interspeech 2025 URGENT Challenge, ranking 2nd in Track 1.