Human Recognition Model Using CNN
Mridula S. Korde, Abhishek Joshi, Aditya Divyakant Shrivastava · INTERNATIONAL JOURNAL OF NEXT-GENERATION COMPUTING · 2022
Since the ever increasing demand of social platforms and social media, it is highly recommended to check age and gender of the person using social websites to avoid any age restricted content. Automatic age and gender classification offers variety of applications majorly authentication, visual surveillance, customer care, biometrics. A convolutional neural network (CNN) is emerging tool for image processing and human identification. In this paper, deep- convolutional neural networks (CNN) is proposed to learn structures of human’s face and hence improve performance for recognition of human’s gender and age significantly. Here, we propose a less complex convolutional neural network design which can be effectively used in case of a small amount of learning data too. The attempts include lowering the number of parameters, boosting the network's depth, and adjusting the dropout rate to have better matching with current state-of-the-art methods using the Adience benchmark.