Automatic Gender Recognition Using Fusion of Facial Strips

Abstract

We propose a fully automatic system that detects and normalizes faces in images and recognizes their genders. To boost the recognition accuracy, we correct the in-plane and out-of-plane rotations of faces, and align faces based on estimated eye positions. To perform gender recognition, a face is first decomposed into several horizontal and vertical strips. Then, a regression function for each strip gives an estimation of the likelihood the strip sample belongs to a specific gender. The likelihoods from all strips are concatenated to form a new feature, based on which a gender classifier gives the final decision. The proposed approach achieved an accuracy of 88.1% in recognizing genders of faces in images collected from the World-Wide Web. For faces in the FERET dataset, our system achieved an accuracy of 98.8%, outperforming all the six state-of-the-art algorithms compared in this paper.

Gender and gaze gesture recognition for human-computer interaction

“…To further validate our method, we compare the accuracy of our method to 8 other state-of-the-art methods with an evaluation of their respective limitations, detailed in Table 1. The accuracy of the proposed method is only slightly lower than [14] and [15] on the FERET database and [13] on the LFW database, but outperforms the others in comparison. However it should be noted that [14] used both fa and fb subsets of the FERET database containing replication of most subjects, while [15] used a large number of classifiers and did not specify the evaluation method.…”

“…one may experience a drop in classification rates as much as 6% [15]. In addition, applying histogram equalisation is a common pre-processing procedure that may increase classification rate.…”

“…They include intensity values from greyscale images [12], LBP [13,14], facial strips [15], Haar-like features [9], SIFT features [16], etc. These features can be extracted globally from complete face images or locally from defined subregions of the face images.…”

“…In addition, only controlled databases are used for training and testing in this research so that the applicability of this approach to dynamic environments remains unevaluated. Similarly, another study [15] employs 10 regression functions to conduct region-based classifications and feed the vector of classification results into an SVM to generate the final decision. Despite the 98.8% accuracy with the FERET database they reported, they didn't illustrate their evaluation method and the split of training and testing data.…”

Improved Gender Recognition during Stepping Activity for Rehab Application Using the Combinatorial Fusion Approach of EMG and HRV

“…To guarantee the accurate signal of a single modality system, the fusion of two or more signals is significant to improve the performance of the system [39,[41][42][43][44][45][46][47][48]. To achieve robust and discriminative performance for gender recognition, a fusion of EMG and HRV [40,41,49] is proposed before feeding into the interface system in order to control the level of MR valve stiffness of the stepper before the rehabilitation exercise can be emulated [40,41].…”