mlsanityData to 14 Oct 2026

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

arXiv 2505.18533 ↗First posted 24 May 2025eess.AS cs.AIcited by 2

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 18 of its 29 references that are in the map. First posted on arXiv 24 May 2025.

Builds on

18 of 29 references.

References · 18

  1. 2021WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processingcited by 1,657
  2. 2022UTMOS: UTokyo-SaruLab System for VoiceMOS Challenge 2022cited by 390
  3. 2022BigVGAN: A Universal Neural Vocoder with Large-Scale Trainingcited by 251
  4. 2021DNSMOS P.835: A Non-Intrusive Perceptual Objective Speech Quality Metric to Evaluate Noise Suppressorscited by 172
  5. 2022Universal Speech Enhancement with Score-based Diffusioncited by 75
  6. 2022TF-GridNet: Integrating Full- and Sub-Band Modeling for Speech Separationcited by 63
  7. 2021Decoupling Magnitude and Phase Estimation with Deep ResUNet for Music Source Separationcited by 48
  8. 2024SpeechBERTScore: Reference-Aware Automatic Evaluation of Speech Generation Leveraging NLP Evaluation Metricscited by 38
  9. 2022VoiceFixer: A Unified Framework for High-Fidelity Speech Restorationcited by 37
  10. 2024ICASSP 2024 Speech Signal Improvement Challengecited by 16
  11. 2024Towards High-Quality and Efficient Speech Bandwidth Extension with Parallel Amplitude and Phase Predictioncited by 14
  12. 2020Channel-wise Subband Input for Better Voice and Accompaniment Separation on High Resolution Musiccited by 13
6 more
  1. 2020Audio Codec Enhancement with Generative Adversarial Networkscited by 12
  2. 2023Toward Universal Speech Enhancement for Diverse Input Conditionscited by 10
  3. 2023ICASSP 2023 Speech Signal Improvement Challengecited by 5
  4. 2024KS-Net: Multi-band joint speech restoration and enhancement network for 2024 ICASSP SSI Challengecited by 5
  5. 2024BS-PLCNet: Band-split Packet Loss Concealment Network with Multi-task Learning Framework and Multi-discriminatorscited by 4
  6. 2024RaD-Net: A Repairing and Denoising Network for Speech Signal Improvementcited by 4

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026GAP-URGENet: A Generative-Predictive Fusion Framework for Universal Speech Enhancementcited by 1
  2. 2026UniPASE: A Generative Model for Universal Speech Enhancement with High Fidelity and Low Hallucinationscited by 0