Wenxuan Wu | CUHK

About Me

I am a PhD candidate from The Chinese University of Hong Kong, under the supervision of Prof. Helen Meng and Prof. Haizhou Li. I am also advised by Prof. Shuai Wang. My current research focuses on speech extraction, generative models, and audio-visual learning. Before that, I obtained my master’s degree in Machine Learning from KTH Royal Institute of Technology through the HIT-KTH 3+2 program, where I finished my master’s thesis on vehicle trajectory prediction with Dr. Yuchao Li.

Research Interests

Speech processing: Speech Extraction, SpeechLLM
Deep learning: Generative Models, Self-supervised Learning

News

[Nov. 2025] We introduce ELEGANCE, incorporating three different linguistic knowledge from LLMs into AV-TSE models: linguistic constraints, linguistic prediction, and linguistic prior.
[July. 2025] We introduce MeMo, an online AVTSE framework equipped with memory banks.
[June. 2025] Reported our work on AV-TSE at the SpeechHome seminar.
[May. 2025] Two papers accepted to Interspeech 2025.
[Mar. 2025] One journal paper accepted to IEEE JSTSP.
[Oct. 2024] One paper accepted to IEEE ICASSP 2025.
[May. 2024] One paper accepted to CVPR Sight and Sound workshop 2024.
[Mar. 2024] One paper accepted to IJCNN 2024.

Publications

arXiv

ELEGANCE: Efficient LLM Guidance for Audio-Visual Target Speech Extraction

Wenxuan Wu, Shuai Wang, Xixin Wu, Helen Meng, Haizhou Li

Submitted.

PDF Project Page

arXiv

MeMo: Attentional Momentum for Real-time Audio-visual Speaker Extraction under Impaired Visual Conditions

Junjie Li*, Wenxuan Wu*, Shuai Wang, Zexu Pan, Kong Aik Lee, Helen Meng, Haizhou Li (*Equal contribution)

Submitted.

PDF Project Page

Interspeech 2025

Incorporating Linguistic Constraints from External Knowledge Source for Audio-Visual Target Speech Extraction

Wenxuan Wu, Shuai Wang, Xixin Wu, Helen Meng, Haizhou Li

Interspeech 2025.

PDF

IEEE JSTSP

C2AV-TSE: Context and Confidence-aware Audio Visual Target Speaker Extraction

Wenxuan Wu, Xueyuan Chen, Shuai Wang, Jiadong Wang, Lingwei Meng, Xixin Wu, Helen Meng, Haizhou Li

IEEE Journal of Selected Topics in Signal Processing.

PDF

CVPR Sight and Sound Workshop 2024

AVHuMAR: Target Speech Extraction with Pre-trained AV-HuBERT and Mask-And-Recover Strategy

Wenxuan Wu, Xueyuan Chen, Xixin Wu, Haizhou Li, Helen Meng

IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshop (CVPRW), 2024.

PDF

Interspeech 2025

DiffDSR: Dysarthric Speech Reconstruction Using Latent Diffusion Model

Xueyuan Chen, Dongchao Yang, Wenxuan Wu, Minglin Wu, Jing Xu, Xixin Wu, Zhiyong Wu, Helen Meng

Interspeech 2025.

PDF

ICASSP 2025

Large Language Model Can Transcribe Speech in Multi-Talker Scenarios with Versatile Instructions

Lingwei Meng, Shujie Hu, Jiawen Kang, Zhaoqing Li, Yuejiao Wang, Wenxuan Wu, Xixin Wu, Xunying Liu, Helen Meng

IEEE International Conference on Acoustics, Speech, and Signal Processing, 2025.

PDF

Awards

Postgraduate Studentships, CUHK
Karl Engvers Foundation Scholarship, KTH
KTH Covid-19 Financial Aid Scholarship, KTH
Study Excellence Scholarship, HIT
Honorable Mention, MCM 2019

Services

IEEE ICASSP 2026, reviewer
IEEE/ACM TASLP, reviewer
IEEE IJCNN 2024, 2025, reviewer
International Doctoral Forum 2023, organizer member, CUHK
Master thesis opponent, KTH

Teaching

ENGG2440 Discrete Mathematics for Engineers, Spring terms, CUHK
ENGG1120 Linear Algebra for Engineers, Fall terms, CUHK
Special Task Force TA, CUHK