Diogenes

摘要

Diogenes is a web image search agent that retrieves web pages and associates a person name with each facial image on those pages. To accomplish its goal, Diogenes relies on two types of evidence: Visual and textual. A face detection module examines the images on the web page for human faces. A face recognition module identifies the face by using a database of known person images. A text/HTML analysis module analyses the body of the text with the aim of finding clues about who appears in each image. The outputs of the face detection, text/HTML analysis and face recognition modules are merged using different evidence combination mechanisms to classify each image. Diogenes have been successfully used to retrieve person images from the web with an average precision surpassing those of some commercial image search engines.