Multi Model Approach to Extract Human Features in Real Time

Face detection and analysis systems has been growing in last few years for various applications. Since the hardware performance increase in last few years, useof Deep Learning, Convolution Neural Network, Face detection, Face analysis techniques is increasing and day by day developed models are breaking accuracies of previous models and research in various tasks. Facial analysis system with age, gender and emotion recognition have been proposed with good accuracies for real-time and non-real time both. The present research paper focuses to provide a robust system architecture for age, gender and emotion recognition in real time which can be use in commercial, healthcare, and many more industries. To achieve this a literature survey is done on the same topic with previous researches to compare their results. The final model architecture proposed in this research paper is efficient and fast and provides accurate results as compare to previous researches