Recognizing faces in news photographs on the web

Abstract

Abstract-We propose a graph based method in order to recognize the faces that appear on the web using a small training set. First, relevant pictures of the desired people are collected by querying the name in a text based search engine in order to construct the data set. Then, detected faces in these photographs are represented using SIFT features extracted from facial features. The similarities of faces are represented in a graph which is then used in random walk with restart algorithm to provide links between faces. Those links are used for recognition by using two different methods.

Automatic Face Annotation in News Images by Mining the Web

“…The application of face recognition techniques to the problem of automatic annotation of news-or of captioned images-has gained large interest in recent years [1], [2], [3], [4], [5], [6]. This problem is particularly challenging because it involves faces captured in broadly varying configurations with respect to viewpoint, expression, illumination and so on.…”

“…Other graph based approaches for annotating faces with names in news have been presented in [5] and [6]. The authors of the former use SIFT features to construct a similarity graph of faces in a small training set; they then annotate unknown faces with the nearest known face node or with the closest category, being a category a set of faces of the same person.…”