Multiple Hierarchical Technique to Predict the Gender of a Person based on 3 Sigma Control Limits on Neural Network

Objective: Gender Classification is one of the most important applications in Artificial Intelligent Systems. The objective of this research work is to classify the human gender based on multiple level decisions. Since many years, a great deal of effort has been made to gender recognition from face images. It is not straightforward to achieve the same accuracy level in real-world environment. The proposed approach can give solution to these problems. Methods: In this paper, multiple levels hierarchical techniques based on 3 Sigma control limits on Neural Network is applied for gender recognition to get the desired objectives. In order to achieve this, the proposed algorithm considers the Artificial Neural Network as basic classifier. Here, in Initial Level Hierarchy, facial features are given as input to the Neural Network. Then, the output represents the gender classification from the Neural Network is extracted. The next level of classification can be done in Core Hierarchical Decision. Findings: This paper provides an effective approach that classifies human gender in computer vision applications. In the proposed research, a Feed Forward Neural Network works at the primary level, based on the outcome of the primary level, the further classification is done in the next higher level hierarchically. In this research, there are 1000 gray-scale with 256 gray levels facial images used for experiment. Each image size is normalized to 64×64. Among the 1000 experimental images, 800 images are used as training data, and the remaining are used as test images. Prediction of the gender is more accurate and effectively achieved the success rate of 95 percent. Applications: The proposed algorithm can play an important role in many computer vision based applications such as human-computer interaction, surveillance, biometrics, demographic studies and targeted advertising.