Naming of faces on photographs by using the relationship between people

Abstract

In this study, we are aiming to name faces of people in photographs which are collected from web. In photographs from the web, some people appear more frequently than some other people. In addition to that, since there are so many examples of more frequently appearing people, naming of these people is simpler than naming of other less frequently appearing people. Therefore, naming of less common people on photographs seems to be problematic. In this study, we are aiming to propose a method which increases the correctness of naming infrequently appearing people. By this purpose, we first try to match the names of most frequently appearing people with their faces. After that, if there are any names remaining, then we assume that these names match with people who are appearing less frequently in the photographs. Therefore, we try to name those people by using the similarity of their faces. For this purpose, we use the politicians' photographs in order to build our dataset, because pair relations can be easily exploited in these photographs. ©2009 IEEE.