8 papers
PhASE-Flow: Phonetic-Conditioned Acoustic Flow Matching in SSL Representation Domain for Speech Enhancement
Jun Gao, Xiaobin Rong, Yu Sun +2
Flow matching (FM) enables high-fidelity generation, while self-supervised learning (SSL) speech models provide hierarchical representations spanning acoustic and phonetic levels.…
HALO: Half-Frame-Rate Adaptive Learnable Operator for Lightweight STFT-Based Speech Enhancement
Jiadong Zhao, Dahan Wang, Yu Sun +5
STFT-based speech enhancement typically adopts overlapping analysis frames. While overlap is essential for stable STFT processing, it makes adjacent frames highly correlated, causi…
Reducing Linguistic Hallucination in LM-Based Speech Enhancement via Noise-Invariant Acoustic-Semantic Distillation
Zheng Wang, Xiaobin Rong, Hang Su +6
Language model (LM)-based speech enhancement (SE) can generate natural-sounding speech, but under severe noise it often suffers from unreliable conditioning, leading to perceptuall…
GAP-URGENet: A Generative-Predictive Fusion Framework for Universal Speech Enhancement
Xiaobin Rong, Yushi Wang, Zheng Wang +1
We introduce GAP-URGENet, a generative-predictive fusion framework developed for Track 1 of the ICASSP 2026 URGENT Challenge. The system integrates a generative branch, which perfo…
StuPASE: Towards Low-Hallucination Studio-Quality Generative Speech Enhancement
Xiaobin Rong, Jun Gao, Zheng Wang +3
Achieving high perceptual quality without hallucination remains a challenge in generative speech enhancement (SE). A representative approach, PASE, is robust to hallucination but h…
PASE: Leveraging the Phonological Prior of WavLM for Low-Hallucination Generative Speech Enhancement
Xiaobin Rong, Qinwen Hu, Mansur Yesilbursa +2
Generative models have shown remarkable performance in speech enhancement (SE), achieving superior perceptual quality over traditional discriminative approaches. However, existing…
A Lightweight Hybrid Dual Channel Speech Enhancement System under Low-SNR Conditions
Zheng Wang, Xiaobin Rong, Yu Sun +3
Although deep learning based multi-channel speech enhancement has achieved significant advancements, its practical deployment is often limited by constrained computational resource…
TS-URGENet: A Three-stage Universal Robust and Generalizable Speech Enhancement Network
Xiaobin Rong, Dahan Wang, Qinwen Hu +3
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, Rob…