Facial age estimation by learning from label distributions

Xin Geng, Kate Amanda Smith-Miles, Zhi-Hua Zhou

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review

56 Citations (Scopus)


One of the main difficulties in facial age estimation is the lack of sufficient training data for many ages. Fortunately, the faces at close ages look similar since aging is a slow and smooth process. Inspired by this observation, in this paper, instead of considering each face image as an example with one label (age), we regard each face image as an example associated with a label distribution. The label distribution covers a number of class labels, representing the degree that each label describes the example. Through this way, in addition to the real age, one face image can also contribute to the learning of its adjacent ages. We propose an algorithm named IIS-LLD for learning from the label distributions, which is an iterative optimization process based on the maximum entropy model. Experimental results show the advantages of IIS-LLD over the traditional learning methods based on single-labeled data.
Original languageEnglish
Title of host publicationProceedings of the Twenty-Fourth AAAI Conference on Artifical Intelligence
EditorsMaria Fox, David Poole
Place of PublicationCalifornia USA
PublisherIEEE, Institute of Electrical and Electronics Engineers
Pages451 - 456
Number of pages6
Publication statusPublished - 2010
EventAAAI Conference on Artificial Intelligence 2010 - Atlanta, United States of America
Duration: 11 Jul 201015 Jul 2010
Conference number: 24th


ConferenceAAAI Conference on Artificial Intelligence 2010
Abbreviated titleAAAI 2010
CountryUnited States of America

Cite this