Multimedia Information Retrieval over WWW
(PhD student opening)

Dr Yongsheng Gao

Email: yongsheng.gao@griffith.edu.au
http://maxwell.me.gu.edu.au/yg

  
With the current ever growing amount of multimedia data over World Wide Web (WWW), searching for information of interest is becoming increasingly difficult and of great commercial potential. The multimedia data over WWW exists in the form of programs that typically consist of a combination of media types such as visual, audio and text. Most of the search engines still only allow search of textual document (such as Yahoo, Netscape and IE), although search and retrieval with images and visual features has been more demanding for quite a while. It is very attractive that image retrieval is not restricted by the user's language and knowledge. The recent MPEG-4 and MPEG-7 standards are addressing the representation and coding of multimedia contents.This research intends to investigate the multimedia search and retrieval over WWW. In the image search and retrieval step, our proposed LHD technique will be fused with existing textual retrieval approaches. The proposed Line Edge Map is a novel compact object coding and representation form. It is less sensitive to illumination changes and demands less storage and computational requirements than edge map.The student is required to understand the above techniques and relevant backgrounds, investigate and improve the implementations, and propose new ideas. C/C++ programming skills and image processing knowledge are required.