Biography

Yu Wang is an Associate Professor at Cooperative Medianet Incorporative Center (CMIC), Shanghai Jiao Tong University (SJTU). Prior to SJTU, he was a Senior Research Associate working in the Machine Intelligence Laboratory, University of Cambridge. He was working with Prof. Mark Gales, Prof. Phil Woodland and Dr. Kate Knill. He obtained his PhD study in Speech Processing in the Speech and Audio Processing Laboratory at Imperial College London, supervised by Mike Brookes. Following his graduation of PhD, he joined University of Cambridge as a Research Associate and started to work as a key member on the Automated Language Teaching and Assessment (ALTA) project (funded by Cambridge Assessment), on which we released an end-to-end deep learning-based automatic spoken language assessment platform Speak&Improve in 2019 and it is now officially suggested as the practising platform for Cambridge Linguskill international English test. Since 2019, he started to work on the Machine Translation for English Retrieval of Information in Any Language (MATERIAL) project (funded by IARPA), on which he was the technical lead for the spoken language processing contribution at CUED. His current research interests focus on Natural Language Processing, Multi-modal Dialogue System and Large Langage Model.

We cordially invite masters and doctoral students, as well as postdoctoral researchers in the fields of computer science and electronic information, to join our research group. Our group is also keen to welcome undergraduate students with a passion for AI research. Here, we offer a vibrant and innovative learning environment where you can explore AI with outstanding peers, learning and growing together. If you are interested in joining us, please contact me.

实验室长期诚邀计算机与电子信息领域的硕士、博士研究生,以及博士后研究人员。同时,我们也非常欢迎对科研有兴趣的本科生加入我们的研究团队。这里提供一个充满活力和创新的学习环境,你可以与优秀的同学们一起探索人工智能,共同学习和成长。如果您有意加入我们,请与我联系

Positions

  • 2020 - Present, Associate Professor, Cooperative Medianet Incorporative Center,
    Shanghai Jiao Tong University
  • 2019 - 2020, Senior Research Associate, Machine Intelligence Laboratory,
    University of Cambridge
  • 2015 - 2019, Research Associate, Machine Intelligence Laboratory,
    University of Cambridge
Interests
  • Natural Language Processing
  • Speech and Language
  • Multimedia Computing
Education
  • Doctor of Philosophy, 2011 – 2015

    Imperial College London, Department of EEE

  • Master of Science (MSc), 2009 – 2010

    Imperial College London, Department of EEE

  • Bachelor of Engineering (BEng), 2005 – 2009

    Huazhong University of Science and Technology, Department of EIE

Experience

 
 
 
 
 
Cooperative Medianet Incorporative Center, Shanghai Jiao Tong University
Associate Professor
Cooperative Medianet Incorporative Center, Shanghai Jiao Tong University
December 2020 – Present
 
 
 
 
 
Machine Intelligence Laboratory, University of Cambridge
Senior Research Associate
Machine Intelligence Laboratory, University of Cambridge
September 2019 – November 2020
 
 
 
 
 
Machine Intelligence Laboratory, University of Cambridge
Research Associate
Machine Intelligence Laboratory, University of Cambridge
August 2015 – September 2019

Projects

*
Large Language Model
Large language models (LLMs) are advanced artificial intelligence systems capable of understanding, generating, and engaging in human-like text-based conversations across a wide range of topics and languages. Our research focuses on: hallucination evaluation and elimination, model compression and medical applications.
Large Language Model
Multimodal dialogue and perception
Multimodal dialogue and perception enhance human-machine interaction by integrating various communication forms and sensory inputs for more natural and intuitive exchanges. Our research focuses on: multi-turn video-grounded dialogue generation, speech perception and understanding and multimodel representation learning.
Multimodal dialogue and perception
Retrieval-Augmented Generation
Retrieval-Augmented Generation (RAG) technology enhances natural language generation by incorporating information retrieved from a large database or documents, thus improving the relevance and accuracy of the generated content. Our research focuses on: cross-modal information retrieval, knowledge selection and knowledge-enhanced dialogue generation.
Retrieval-Augmented Generation

Recent Publications (2023 Onwards)

Quickly discover relevant content by filtering publications.
* indicates corresponding authors
(2024). MING-MOE: Enhancing Medical Multi-Task Learning in Large Language Models with Sparse Mixture of Low-Rank Adapter Experts. arXiv preprint arXiv:2404.09027.

Cite Project

(2024). Automatic Interactive Evaluation for Large Language Models with State Aware Patient Simulator. arXiv preprint arXiv:2403.08495.

Cite Project

(2024). CE-VDG: Counterfactual Entropy-based Bias Reduction for Video-grounded Dialogue Generation. COLING 2024.

Cite

(2024). DictLLM: Harnessing Key-Value Data Structures with Large Language Models for Enhanced Medical Diagnostics. arXiv preprint arXiv:2402.11481.

Cite Project

(2024). Leveraging Diverse Modeling Contexts with Collaborating Learning for Neural Machine Translation. IEEE/ACM Transactions on Audio, Speech, and Language Processing.

Cite

Teaching

 
 
 
 
 
程序设计(荣誉)
人工智能专业 80学时 4学分
September 2022 – Present
 
 
 
 
 
程序设计思想与方法(C++)
工科平台 80学时 4学分
September 2021 – January 2022

Contact