78838 Engineering & Technology /_upload/column/33/b8/78776/picture.jpg /_s1134/Engineering/list.psp /_s1134/Engineering/main.psp homepage  Projects  Engineering & Technology
2024-12-18 Bruce Xinbo Yu Human behavior understanding, Multi-agent AI, efficient fine-tuning /_upload/article/images/3c/3e/ba159192401e9d1df51ce07daa9b/3c6157da-fa3b-4a52-841d-9fd2c5fc5dce.jpg


Assistant Professor,Zhejiang University


Lab Name:

Emerging AI Lab


Homepage

https://bruceyo.github.io


Email

xinboyu@intl.zju.edu.cn


Research

Human behavior understanding, Multi-agent AI, efficient fine-tuning


Number of Participants

1-4 Students


Biography

Dr. Bruce X.B. Yu earned his Ph.D. from The Hong Kong Polytechnic University. His primary research expertise focuses on vision-based human behavior understanding. His works have been published on top venues such as: TPAMI, ACM CSUR, Pattern Recognition, ICCV, AAAI, IJCAI, etc. His research outcomes also get international recognition such as best paper and best presentation awards. He has been actively conducting interdisciplinary in research topics such as driver behavior analysis, business intelligence, surgical robotics, etc. Besides application-driven research, he also works on fundamental research problems such as LLM-driven reasoning, multi-agent AI, and efficient fine-tuning.


What You Can Expect in the Project

Based on your research or major background, you will select a topic related to my research expertise, which can be both applied or theoretical research around human behavior understanding. The exact applied research topic can be human action detection, anomaly prediction, etc. While, theoretical research can be domain generalization, domain shift, data efficiency, etc. Techniques or tools used to tackle the problem can be very open, such as LangGraph, AutoGen, Crew AI, etc. During the internship, you will gain useful research experience in AI such as: data collection, building a model, training or fine-tuning a model, writing research papers, presenting research outcome, etc.


Application Specific Requirements

  • The project will mainly focus on driver behavior analysis using sensors such as EEG, eye tracking, cameras, etc. Students applying for this project should have strong programming experience using tools such as PyTorch, LangChain, GPT API calls, etc.


Apply