Abstract:
Automatic human gender and age recognition has started catching the attention of
researchers due to its possible wide application pool. To name a few it can be used in
Human-Computer Interaction (HCI) systems to tune the context appropriately to suit
the target person's gender and age, to monitor specific gender or age restricted areas
in surveillance systems, to make targeted advertising where the relevant
advertisement/information to the audience can be channelled from electronic
billboard systems, in automated biometric data acquisition and for content based
search in which identifying the gender and age reduces the search space significantly.
This project focuses on the area of face processing and aims at designing a reliable
framework to facilitate face, gender, and gender group recognition. A framework has
been optimized for the task of face cropping, gender, and age group recognition. It
makes an extensive experiment with row pixel intensity valued features and Discrete
Cosine Transform (DCT) coefficient features with Principal Component Analysis
and k-Nearest Neighbour classification to identify the best recognition approach. The
final results show approaches using DCT coefficient outperform their counter parts
resulting in a 99% correct gender recognition rate and 68% correct age group
recognition rate (considering four distinct age groups) in unseen test images.
Detailed experimental settings and obtained results are clearly presented and
explained in this report.
