No abstract

This paper presents a thorough study of gender classification methodologies performing on neutral, expressive and partially occluded faces, when they are used in all possible arrangements of training and testing roles. A comprehensive comparison of two representation approaches (global and local), three types of features (grey levels, PCA and LBP), three classifiers (1-NN, PCA+LDA and SVM) and two performance measures (CCR and d ′ ) is provided over single-and cross-database experiments. Experiments revealed some interesting findings, which were supported by three non-parametric statistical tests: when training and test sets contain different types of faces, local models using the 1-NN rule outperform global approaches, even those using SVM classifiers; however, with the same type of faces, even if the acquisition conditions are diverse, the statistical tests could not reject the null hypothesis of equal performance of global SVMs and local 1-NNs.

For more information about UCLan's research in this area go to http://www.uclan.ac.uk/researchgroups/ and search for . For information about Research generally at UCLan please go to http://www.uclan.ac.uk/research/ All outputs in CLoK are protected by Intellectual Property Rights law, including CLoK Central Lancashire online Knowledge www.clok.uclan.ac.uk

Abstract. This paper evaluates the discriminant capabilities of face parts in gender recognition. Given the image of a face, a number of subimages containing the eyes, nose, mouth, chin, right eye, internal face (eyes, nose, mouth, chin), external face (hair, ears, contour) and the full face are extracted and represented as appearance-based data vectors. A greater number of face parts from two databases of face images (instead of only one) were considered with respect to previous related works, along with several classification rules. Experiments proved that single face parts offer enough information to allow discrimination between genders with recognition rates that can reach 86%, while classifiers based on the joint contribution of internal parts can achieve rates above 90%. The best result using the full face was similar to those reported in general papers of gender recognition (>95%). A high degree of correlation was found among classifiers as regards their capacity to measure the relevance of face parts, but results were strongly dependent on the composition of the database. Finally, an evaluation of the complementarity between discriminant information from pairs of face parts reveals a high potential to define effective combinations of classifiers.

Abstract. This paper proposes a gender recognition scheme focused on local appearance-based features to describe the top half of the face. Due to the fact that only the top half of the face is used, this is a feasible approach in those situations where the bottom half is hidden. In the experiments, several face detection methods with different precision levels are used in order to prove the robustness of the scheme with respect to variations in the accuracy level of the face detection process.