徐东教授

Title

Towards Impactful Research:From Visual Domain Adaptation to Deep Video Compression.

abstracts

In this talk, I will first introduce our previous domain adaptationworks, including our pioneering works in developing new domain adaptation(transfer learning) methods for video event recognition, and a series ofsubsequent works for single source domain adaptation, multi-domain adaptation,heterogeneous domain adaptation, domain generalization and deep domainadaptation, as well as their applications in various computer vision tasks.Then I will describe our previous deep video compression works, including thefirst end-to-end optimized deep video compression (DVC) framework, and oursubsequent works including the feature-space video coding (FVC) network, aswell as our recent works for coding mode prediction and stereo videocompression.

Introduction

Prof. Dong Xu is a Professor in the Computer Science Department, TheUniversity of Hong Kong, where he serves as the Director of the JC STEM Lab ofMultimedia and Machine Learning. After receiving his PhD degree from Universityof Science and Technology of China in 2005, he worked as a postdoctoralresearch scientist at Columbia University, a tenure-track and tenured facultymember at Nanyang Technological University, and the Chair in ComputerEngineering at The University of Sydney.

Prof. Xu is an active researcher in the areas of computer vision,multimedia and machine learning. He was selected as a Clarivate AnalyticsHighly Cited Researcher twice in 2021 and 2018. He was also selected as anAustralian Research Council Future Fellow (Level 3, Professorial Level) in 2018and awarded the IEEE Computational Intelligence Society Outstanding EarlyCareer Award in 2017. He has published more than 200 papers in IEEETransactions and leading conferences including CVPR, ICCV, ECCV, ACM MM, ICML and NeurIPS. His co-authored works (with his former PhDstudents) received the Best Student Paper Award in CVPR 2010 and the IEEETransactions on Multimedia Prize Paper Award in 2014.

He is/was on the editorial boards of ACM Computing Surveys (SeniorAssociate Editor since October 2022), IEEE Transactions including T-PAMI, T-IP,T-NNLS, T-CSVT and T-MM, and other five journals, as well as served as a guesteditor of more than ten special issues in multiple journals (e.g., IJCV,IEEE/ACM Transactions). He served as the Program Coordinator of ACM Multimedia2024, a steering committee member of ICME (2016-2017) and a Program Co-chair offive international conferences/workshops (e.g., ICME 2014). He was alsoinvolved in the organization committees of many international conferences andserved as an area chair of leading conferences such as ICCV, CVPR, ECCV, ACM MMand AAAI. He received the Best Associate Editor Award of T-CSVT in 2017. He isa Fellow of IEEE and IAPR (The International Association for PatternRecognition).

徐东教授

题目

迈向有影响力的研究：漫谈视觉域适应和深度视频压缩

摘要

在本次讲座中，徐教授将首先介绍其团队在视觉域适应方面的工作，包括首次将域适应(Domain Adaptation)/ 迁移学习引入到计算机视觉领域中视频事件识别任务的工作，以及后续在单源域适应、多源域适应、异构域适应、域泛化和深度域适应等方向的工作。然后，徐教授将介绍其团队在深度视频压缩方面的工作，包括首个基于端到端优化的深度视频压缩框架DVC和后续的工作，包括特征空间视频编码网络FVC以及最近在编码模式预测和双目视频压缩等方面的工作。

报告人简介

徐东教授于2001年和2005年在中国科学技术大学取得学士和博士学位,目前担任香港大学计算机系教授。徐东教授曾在微软亚洲研究院、香港中文大学和哥伦比亚大学从事研究工作，并在南洋理工大学和悉尼大学任教。徐东教授在计算机视觉、多媒体信号处理以及机器学习等领域做出了重要贡献,在IEEE Transactions和国际顶级会议上发表了200余篇论文，其中两篇论文分别获得T-MM 2014最佳论文奖和CVPR 2010最佳学生论文奖。徐东教授是IEEE和IAPR Fellow，于2018和2021年两次入选科睿唯安 (Clarivate Analytics)高被引学者。

徐东教授目前担任ACM Computing Surveys (CSUR)资深副主编，曾担任T-PAMI，T-IP，T-NNLS,T-MM和T-CSVT等学术期刊副主编，以及10余个国际期刊 (比如IJCV)的客座编委，曾担任ACMMM 2024 Program Coordinator，VALSE和ICME指导委员会成员，五个国际会议 (比如ICME 2014)的程序委员会协同主席，以及AAAI/ ICCV/ ECCV/CVPR等国际会议区域主席。