Mapping between image regions and caption concepts of captioned depictive photographs
Rowe, Neil C.
MetadataShow full item record
We discuss the obstacles to inference of correspondances between objects within photographic images and their counterpart concepts in descriptive captions of those images. This is important for information retrieval of photographic data since its content analysis is much harder than linguistic analysis of its captions. We argue that the key mapping is between certain caption concepts representing the "linguistic focus" and certain image regions representing the "visual focus". The mapping is one-to-many, however, and many image regions and captions concepts are not mapped at all. We discuss some domain-independent constraints that can restrict potential mappings. We also report on experiments testing our criteria for visual focus of images.
This paper appeared in the AAAI-98 Workshop on Representations for Multi-Modal Human-Computer Interaction, July 1998, Madison, Wisconsin, USA.
Showing items related by title, author, creator and subject.
Carpenter, Daniel K.; Ebner, Donald J. (Monterey, California. Naval Postgraduate School, 1993-12);This thesis is an evaluation of current computer software applications and their use in strategic planning. Military and business strategy are reviewed and discussed and the premise for sharing ideas between communities ...
Gupton, Kevin; Abbott, Jeff; Blais, Curtis; Diallo, Saikou Y.; Heffner, Kevin; Turnitsa, Chuck (2011);Standard information exchange data models (IEDMs), such as the Joint Consultation Command and Control IEDM (JC3IEDM) managed by the Multilateral Interoperability Programme (MIP) and the National Information Exchange Model ...
Stewart, James Jeremiah. (Monterey, California. U.S. Naval Postgraduate School, 1967);Various methods are studied for the relocation, or movement, including address mapping, of programs within a multiprogrammed digital computer. The aim of doing so is to determine the best method for use in the limited ...