9月21日学术报告(微软中国研究院 Dr Xian-Sheng Hua
类别：学术报告 发布人：admin 浏览次数： 次 发布时间：2015-09-17 15:10
题目：From Big Multimedia Data to Real-World Visual Search(& an introduction to the upcoming Alibaba Image Search Challenge)

时间：2015年9月21日上午10:00
地点：武汉大学本科生院楼（原出版社大楼）南613会议室
报告人：Dr Xian-Sheng Hua

报告人国籍：中国

报告人简介：2015年加入阿里巴巴搜索事业部，任研究员/资深总监，负责大数据多媒体内容分析及搜索算法团队。在此之前先后在微软中国研究院、微软总部Bing搜索引擎以及微软美国研究院任主管研究员、首席研发主管以及资深研究员。北京大学博士毕业，一直从事图像、视频内容分析和搜索方面的研究和开发工作。在国际会议和期刊上发表论文250余篇，拥有50余项授权专利，H Index为50。2008年因视频搜索方面的突出贡献获麻省理工学院《技术评论》TR35“全球35个35岁以下青年创新者”称号，2012年任多媒体领域顶级国际会议ACM Multimeida程序委员会主席，2012和2013年连续两年任IEEE多媒体会议ICME程序委员会主席，还是多个多媒体、AI及搜索领域国际会议的领域主席（包括KDD，SIGIR，IJCAI，MM，ICME等），曾任或现任IEEE Trans. On Multimedia及ACM Trans. On Intelligent Systems and Technology杂志副编，曾获包括ACM Multimedia和IEEE Trans. On CSVT等在内的多项最佳论文奖。

 报告摘要：Visual search has been studied for decades since 1990s and it wascalled “content-based image retrieval”  in the  beginning. It almost died in theearly 21st century and then becomes very hot in recent years mainly due to therapid development
 of distributed computing, high-dimensional indexing, deeplearning and big data analytical techniques. Many visual search  researchprototypes and preliminary products are available to the public. However, isreal-world visual search really coming?  Have we solved all the big technicaland non-technical challenges? Are semantic gaps and intention gaps still there,and do they  really matter? How big data analyses and deep learning techniquescan help here? Are people really using it and why do they
want to use it? Whichdirection that visual search is going towards? What are still missing? In thistalk, we will discuss all these
 questions based on a real-world visual searchengine that has a considerable number of users. And we will also  introduce the upcoming  “Alibaba Large- Scale Image Search Challenge”, in which we will awardmore than 400K RMB  in total to winning teams.

邀请人：胡瑞敏教授   
上一篇：9月17日学术报告(新西兰奥克兰理工大学 Prof. Re
下一篇：10月14日Alex X. Liu教授学术报告