April 11-14, 2007
San Francisco, California

Demonstrations: Description

Machine Assisted Tagging of Images

James Wang, The Pennsylvania State University, USA

Automated annotation of digital pictures has been a highly challenging problem for computer scientists since the invention of computers. The need for computer assisted tagging of images has grown in the museum and archive community as the size digital image collections grows rapidly. In our work, by advancing statistical modeling and optimization techniques, we can train computers about hundreds of semantic concepts using example pictures from each concept. The ALIPR (Automatic Linguistic Indexing of Pictures - Real Time) system of fully automatic and high speed annotation for online pictures has been constructed. Thousands of pictures from an Internet photo-sharing site, unrelated to the source of those pictures used in the training process, have been tested. In this demonstration, we show that a single computer processor can suggest annotation terms in real time with good accuracy.

Demonstration: Demonstrations - 2 [Close-Up]

Keywords: Automatic image annotation, tagging, content based image retrieval, image analysis