Named Faces: putting names to faces

To provide automatic labeling of faces in video, the author has developed Named Faces, a fully functional automated system that builds a large database of name-face association pairs from broadcast news. This article describes how the system detects and recognizes superimposed text in the video, then verifies or repairs the text by comparing it with a large list of automatically generated names found in news stories. Faces found in the video where superimposed names were recognized are tracked, extracted, and associated with the superimposed text. With Named Faces, users can submit queries to find names for faces in video images.