Abstract
This is an extended version of our paper [1] with more results and more contributors in our study, Gender identification has immense value in science and technology. It helps optimizing a software application pertaining to users' gender. In this study, we show how the 3D touch sensor can be used for gender identification. The 3D touch sensor is newly introduced in the handheld mobile devices. This sensor measures how much force the user applies to the touch screen. Therefore, these biometric measurements can be exploited for gender identification. This study develops a special iOS application to collect 3D touch sensor's data from a user. Discrete wavelet transform based features extracted from such data can then be used to train ensemble bagged tree classifier for the purpose of gender identification. The developed iOS application and proposed classifier are tested using data collected from 185 males and 181 females. The results are evaluated in terms of sensitivity and specificity, giving averaged values of 78.2% and 79.1% respectively; however, the sensitivity/specificity could exceed 97% in some cases. Therefore, a 3D touch sensor has a potential value in providing biometric data useful for gender identification.