Yuxiang Mei
ASR · Speech-LLM
About me
I am an M.Eng. candidate in Communication Engineering at Shanghai Normal University, supervised by Prof. Yanhua Long. My research focuses on automatic speech recognition (ASR) and speech large language models (Speech-LLMs), with particular interests in multilingual recognition, parameter-efficient adaptation, and target-speaker ASR.
I am currently an intern at Baidu, focusing on target-speaker automatic speech recognition (TS-ASR).
Selected publications & preprints
Journal & conference papers
TASLP 2026
Zipper-LoRA: Dynamic Parameter Decoupling for Speech-LLM based Multilingual Speech Recognition
Yuxiang Mei, Delai Qiu, Shengping Liu, Jiaen Liang, and Yanhua Long. IEEE Transactions on Audio, Speech, and Language Processing, 2026.
Parameter-efficient multilingual ASR adaptation designed to reduce cross-language interference, with evaluation across 12 languages.
ICASSP 2026
Bridging the Gap: A Comparative Exploration of Speech-LLM and End-to-End Architecture for Multilingual Conversational ASR
Yuxiang Mei, Dongxing Xu, Jiaen Liang, and Yanhua Long. IEEE International Conference on Acoustics, Speech, and Signal Processing, pp. 18797–18801, 2026.
A comparison of Speech-LLM and end-to-end architectures, exploring dual-encoder cross-attention fusion and speech-encoder pre-tuning.
MLC-SLM 2025
SHNU Multilingual Conversational Speech Recognition System for INTERSPEECH 2025 MLC-SLM Challenge
Yuxiang Mei, Yuang Zheng, Dongxing Xu, and Yanhua Long. Workshop on Multilingual Conversational Speech Language Model, pp. 38–42, 2025.
A Whisper + mHuBERT-147 dual-encoder system with Qwen2.5-7B, three-stage training, and language-specific prompting. Ranked 7th out of 25 teams without extra data.
Preprint
PREPRINT · 2026
Unified Target-Speaker ASR with Text and Enrollment Speech Cues
Yuxiang Mei, Yuchen Yan, Dongxing Xu, Jiaen Liang, and Yanhua Long. arXiv preprint, 2026. Submitted to ICLR 2027.
A unified target-speaker ASR framework supporting text-only, enrollment-speech-only, and dual-cue inference within a single model.
Research experience
Baidu · ACG May 2026–Present
AI Speech Algorithm R&D Intern · Shanghai, China
Research on text-guided target-speaker ASR for far-field voice interaction, and a unified framework combining text and enrollment-speech cues.
Unisound · AI Labs Oct 2025–Mar 2026
Speech Foundation Model Research Intern · Xiamen, China
Developed Zipper-LoRA for parameter-efficient multilingual ASR adaptation and completed training, evaluation, and analysis across 12 languages. Published in IEEE TASLP 2026.
Education
Shanghai Normal University Sep 2024–Jun 2027 (expected)
M.Eng. in Communication Engineering · Supervisor: Prof. Yanhua Long
Xi’an Peihua University Sep 2020–Jun 2024
B.S. in Software Engineering
Honors & awards
- National Scholarship for Graduate Students, Ministry of Education of China · 2026
- First-Class Graduate Academic Scholarship, Shanghai Normal University · 2026
- Third-Class Graduate Academic Scholarship, Shanghai Normal University · 2024, 2025
Curriculum vitae
Last updated: October 2026
