您的当前位置: 美高梅4858线路登录  >  科学研究  >  学术动态  >  正文

学术动态

2024国际产学研用合作会议 (长春) 吉林大学人工智能创新应用分论坛海外专场暨计算机科学技术专家讲座(十九)——吴波

发布日期:2024-09-25 发布人: 点击量:

报告题目:Learning and Acquiring Multimodal Commonsense Reasoning

报告时间:2024928 1110

报告方式:腾讯会议

码:172-510-716

人:吴波博士 MIT-IBM Watson AI Lab


报告人简介:

Bo Wu is currently a Researcher with MIT-IBM Watson AI Lab, Cambridge, MA. Bo received his Ph.D. in Computer Science from the Institute of Computing Technology, Chinese Academy of Sciences, Beijing, and he was a Research Scientist at Columbia University, New York City, NY. His research interests encompass deep learning, multimodal learning, computer vision, and natural language understanding, with a focus on visual, linguistic, and user behavior analysis, forecasting, and reasoning. He won the research awards, including the IBM Master Inventor Award, IBM Level-A Accomplishment Award, ACL Best Demo Paper Award, the ACM Turing 50th Student Scholarship, and the SLB Ph.D. Award. His team has excelled in global competitions, securing top positions in the NIST TAC SM-KBP (1st), the ICIP Prediction Challenge (1st), and the Alibaba Global Vision AI Challenge (3rd, top 0.1%), etc. He has organized events such as the ACM MM Grand Challenge SMP, and the CVPR Workshop and Challenge MVCS and MMFM. He has also served in key roles as Area Chair, Track Chair, Senior Program Committee Member, and Program Committee Board Member for the conferences including ACM Multimedia, AAAI, IJCAI, etc.


报告内容简介:

The lecture explores the journey from basic perceptual processes to complex, real-world reasoning tasks. We begin by examining how the human brain and modern artificial systems perceive and understand visual information. Then, we dive into visual cognition, the mental processes that enable recognition, attention, and categorization. The talk progresses to visual reasoning, where we will discuss recent breakthroughs, including advances in several mainstream methods. Applications of these advancements are wide-ranging, and we will demonstrate real use cases where systems make complex decisions based on real-time visual data, highlighting their impact on fields.

主办单位:吉林大学美高梅4858线路登录(中国)App Store

吉林大学软件学院

吉林大学计算机科学技术研究所

符号计算与知识工程教育部重点实验室

仿真技术教育部重点实验室

网络技术及应用软件教育部工程研究中心

吉林大学国家级计算机实验教学示范中心