Human Age and Gender Prediction Using Deep Multi-Task Convolutional Neural Network

Abstract

Gender and age prediction are the key areas of research in the biometric as well as human face recognition applications aimed at effective future prediction and the knowledge discovery about the specific person. The process makes use of assorted approaches and algorithms whereby the deep learning is also the prime in usage patterns. Our research presents a new idea based on modifying the deep network structure and using learning methods of the two other researchers. We made some modification on the structure of the convolutional neural network (CNN) that was used by the first researcher, then, we used two learning methods, which were adopted by the second researcher, Single-Task Learning (STL) and Deep Multi-Task Learning (DMTL) approach, and we present new structure of CNN according to the above two modifications, implemented and evaluated, and the results show the effective performance of our proposed structure. The proposed net presented in this work has the association of Adience-Face Benchmark face dataset, samples of a test and training set, the implementation is performed by Python software.

A Lightweight Deep Convolutional Neural Network Model for Real-Time Age and Gender Prediction

“…Accuracy of 96.02% and 80.64% for gender prediction was obtained from these two datasets and accuracy of 44.36% for age prediction was achieved from the proposed model. Another research was done by two learning methods -single task learning (STL) and deep multi-task learning (DMTL) in [3]. Gained accuracy for CNN+STL and CNN+DMTL is 80.11% and 91.34% for gender prediction.…”