博士论坛

博士论坛第七十四期学术报告信息

作者:admin 日期:2014-07-26 13:26:00 浏览:

报告题目:Large-scale Visual Search

报告人:德克萨斯大学 田奇教授

报告简介:

With the emergence of massive   social multimedia data and prevalence of mobile visual search applications,   techniques towards large-scale visual search and recognition are desired.   Recent decade has witnessed the fast advance of large-scale image search,   thanks to introduction of bag-of-visual-words model based on local invariant   features and the scalable index structure. Generally, an image search system   is involved with several key modules, including feature extraction, visual   codebook learning, feature quantization, index strategy, scoring scheme, and   post processing. Besides, post-processing techniques, such as geometric   verification, query expansion and multi-modal fusion, can be plugged in to   boost the retrieval performance.

In the first part of the talk, I   will make an overview of large-scale visual search and discuss two   milestones. In the second part, I will figure out four key problems on visual   feature representation, feature quantization, image indexing, and   post-processing in the framework of image search. To address those problems,   we have conducted comprehensive work. I will introduce our recent   representative works and show the related demos. On feature representation,   we have developed a set of binary features, including USB, COGE, etc, and   designed a new regional feature. On feature quantization, I will introduce   our work on codebook learning and efficient quantization. On image indexing,   a series of our recently proposed indexing schemes will be introduced, such   as cross-index, coupled multi-index, super-image index, cascade   category-aware index. Last but not least, on post-processing, I will   introduce our work on geometric verification and query expansion. In the   third part, I will make a summarization and discuss the potential   applications.

报告人简介:

Qi Tian is   currently a Full Professor in the Department of Computer Science, the   University of Texas at San Antonio (UTSA). During 2008-2009, he took one-year   Faculty Leave at Microsoft Research Asia (MSRA) in the Media Computing Group.   He received his Ph.D. in ECE from University of Illinois at Urbana-Champaign   (UIUC) in 2002 and his B.E and M.S degrees from Tsinghua University and   Drexel University in 1992 and 1996, respectively, all from electronic   engineering. Dr. Tian’s research interests focus on multimedia information   retrieval and computer vision and published over 240 refereed journal and   conference papers. He received the Best Paper Award in PCM 2013, ACM ICIMCS   2012 and MMM 2013, a Top 10% Paper Award in MMSP 2011, the Best Student Paper   Award in ICASSP 2006, and was a co-author of a Best Paper Candidate in PCM   2007. His research projects are funded by NSF, ARO, DHS, Google, FXPAL, NEC,   SALSI, CIAS, Akiira Media Systems, HP and UTSA. He received 2010 ACM Service   Award. He is the Guest Editors of IEEE Transactions on Multimedia, Journal of   Computer Vision and Image Understanding, etc, and is the Associate Editor of   IEEE Transactions on Multimedia, IEEE Transactions on Circuits and Systems   for Video Technology (TCSVT) and in the Editorial Board of Journal of   Multimedia (JMM) and Journal of Machine Vision and Applications (MVA).    He is the Guest or Adjunct Professor in Institute of Computing Technology,   Chinese Academy of Science, Xi’an Jiaotong University, USTC, Zhejing   University, Xidian University and a Chaired Professor in Tsinghua University.

时间:7月24日(星期四)下午4:00 

地点:武汉大学国家多媒体软件工程技术研究中心4楼会议室