Gender Classification Using Neck Features Extracted From Profile Images

Vol.3  No.2

Year: 2014

Issue : Mar-May

Title : Gender Classification Using Neck Features Extracted From Profile Images

Author Name : Munehiro Nakamura, Kengo Iwata , Masatoshi Kimura , Haruhiko Kimura

Synopsis :

Gender classification has been one of the emerging issues in the field of security due to increase of women-only floors. For security purpose, this paper presents two biological features extracted from the neck region in a profile image. One is extracted the bump of the neck formed by the laryngeal prominence. Another is the width of the neck that tends to be wider in males than in females. Evaluation experiments for the proposed two features were performed on 50 male and 39 female profile images. The experimental result shows that the proposed method achieved over 95.0% accuracy for both of the male and female images, which overcomes a state-of-the-art method based on Local Binary Patterns for gender classification.

For more details:

imanagerpublications.com