|
Organizing Committee: Tanveer Syeda-Mahmood , Chair Ramesh Jain Linda Shapiro Shih-Fu Chang Program Committee: Ingemar Cox Alberto del Bimbo Takeo Kanade Amarnath Gupta Anil Jain Tom Huang B.S. Manjunath Rakesh Mohan Raymond Ng Rosalind Picard Harpreet Sawhney Michael Swain Murat Tekalp HongJiang Zhang Advance Program Registration |
DESCRIPTIONResearch in computer vision, image processing, and databases is converging towards handling multimedia content in databases. Even though it is acknowledged that accessing based on image content can be a powerful alternative/addition to conventional text annotation-based retrieval, such access has yet to reach the robustness and computational effectiveness of text-based retrieval.The fundamental issues in the design of multimedia databases revolve around fast and robust selection of multimedia data containing an answer to a visual query. The performance of image-content-based access is affected by issues that impact the design of the entire multimedia database system such as multimedia feature extraction, data representation, organization, query formulation and search. In the context of image and video databases, this means research is needed to extract both generic and query-specific representations of image and video data exposing features such as color, texture, motion, and other geometric and text attributes. Effective ways of organizing image and video data representations is needed to enable fast search of the databases. Methods are needed that map the needs of the user's query to a form suitable for indexing the image and video content. Search methods are needed that can localize queries within segmented or unsegmented images/video. New query matching methods that allow for variations in shape, perceptual, semantic and other measures of similarity become increasingly relevant. While automatic methods of database creation are desirable, human intervention is preferred in acquiring incremental specifications of a query and in providing user feedback to combine semantic, perceptual and other criteria in evaluating the results of a query. Finally, new database design frameworks are needed that combine visual information with more traditional database information to respond to queries. Lastly, realistic evaluation criteria are needed for testing the efficiency and accuracy of content-based access on image and video databases of realistic size in domains of interest.
PURPOSEThe purpose of this workshop is to bring together researchers in computer vision, image processing, database and other communities, addressing the above-mentioned issues towards developing robust and computationally effective content-based search of multimedia databases. Original papers are solicited that address the following topics for image and video data:
|
