کد مقاله | کد نشریه | سال انتشار | مقاله انگلیسی | نسخه تمام متن |
---|---|---|---|---|
530313 | 869756 | 2012 | 13 صفحه PDF | دانلود رایگان |

Automatically naming faces in online social networks enables us to search for photos and build user face models. We consider two common weakly supervised settings where: (1) users are linked to photos, not to faces and (2) photos are not labeled but part of a user's album. The focus is on algorithms that scale up to an entire online social network. We extensively evaluate different graph-based strategies to label faces in both settings and consider dependencies. We achieve results on a par with a recent multi-person approach, but with 60 times less computation time on a set of 300K weakly labeled faces and 1.4 M faces in user albums. A subset of the faces can be labeled with a speed-up of over three orders of magnitude.
► We introduce face naming strategies for two types of weakly labeled faces.
► We focus on strategies that scale well to large numbers of users and faces.
► We introduce a face dataset with 1.7 M faces obtained from an online social network.
► We thoroughly investigate the factors that influence the face naming results.
► We achieve state-of-the-art performance with a significantly lower computation time.
Journal: Pattern Recognition - Volume 45, Issue 6, June 2012, Pages 2335–2347