A real-time multi view gait-based automatic gender classification system using kinect sensor

Gender classification plays an important role in many applications such as security and medical applications. Human gender can be classified using different biometric techniques such as face recognition, voice recognition, activity recognition and gait recognition. Different approaches based on gait-recognition have been proposed for the identification of gender. However, performance and accuracy of such systems suffer from the recurring and inherent issues like occlusion of body parts, computational costs and false recognition of 3D joints. The problems can be subdued with deep feature-based analysis and extensive calculation but that may further degrade performance of the system. In this paper, we propose a limited feature-based, Three Dimensional (3D), real time, and multi-view gait-based automatic gender classification system using Microsoft kinect (MS Kinect). A statistical model is molded from the binary logistic regression of the gait data extracted at run time using the sensor. The proposed method is successfully implemented and evaluated by 80 (50 male and 30 female) users. The achieved accuracy rate (97.50%) proves applicability of the model.