演讲人: 微软亚洲研究院资深研究员 张磊 博士

讲座时间: 2010年5月18日（周二）晚6：00~7：30

讲座地点: 伟德国际1946源自英国四层报告厅

讲座内容:

演讲人：微软亚洲研究院资深研究员 张磊 博士

Lei Zhang (Lead Researcher)

Lei Zhang is a lead researcher in the Web Search & Mining Group at Microsoft Research Asia. He currently directs a team pursuing new research directions on social media search. Team projects include user-created content analysis, query understanding, and web-scale image annotation. Lei joined Microsoft Research Asia in 2001 and worked with the Media Computing Group on key projects such as image classification, red-eye detection, face detection and annotation. Three years ago, he moved to his current post in the Web Search & Mining Group, and has worked on large-scale systems related to search and multimedia. Lei earned his B.S. and M.S. in Computer Science from Tsinghua University, in 1993 and 1995. After two years working in industry, He later returned to Tsinghua University, and was awarded a doctorate in Computer Science in 2001. Lei is an IEEE and ACM member, and has served on international conference program committees, including ACM Multimedia, WWW, SIGIR, CIVR, ICME, IJCAI-MIR, ACM Multimedia-CARPE, PCM, MMM, and AIRS. He is the author or co-author of more than 50 published papers in fields such as content-based image retrieval, computer vision, Web search and information retrieval. He also holds five U.S. patents for his innovation in face-detection, red-eye reduction and image retrieval technologies.

演讲摘要：

Despite the great success of Web search engine, Web image search is still in a preliminary stage, mainly depending on text-based search technologies, no matter how many years content-based image retrieval has been studied. In this lecture, we will first go through the history of image search and show the key difficulties in each stage. Then we introduce basic techniques studied in the traditional content-based image retrieval, including feature extraction, relevance feedback and image annotation. Finally, we attempt to discuss the key challenges in Web image search, including how to improve search relevance, how to improve search result quality, and how to improve search result organization. Accordingly, the lecturer will introduce several interesting works recently done in MSRA.