"Unsupervised face identification in TV content using audio-visual sources."

Meriem Bendris et al. (2013)
a service of Schloss Dagstuhl - Leibniz Center for Informatics