Age and Gender Prediction using Gated Residual Attention Network

Abstract

While a great deal of researchers has addressed the challenge of determining the age and sex from face images, these have received significantly fewer spotlights than some other challenges that are particularly linked with face recognition. Relatively to certain other face recognition concerns, the level of performance in this area has not increased significantly. Despite the progress made in age and gender prediction, several challenges like Facial variability, Data bias persist. Each language around the world includes its own lexicons and grammar standards which are aimed at helping individuals of various generations to interact. The decision to adopt is dependent upon the human capability to quickly identify specific characteristics like age and gender from the individual external appearance. This study assumed that AI technologies become even more widespread in variety of sectors and its decision-making abilities might mimic human mind. Besides, this study has created a GRA NET computational intelligence algorithm to Figure out the person’s age and sex.The residual attention network has already been improved and enhanced as well as the infrastructure now encompasses the principle of gate. Age forecasting is a regression issue whereas gender identity is a binary classification issue. Investigations have been done on the publicly available UTK Face dataset. The results obtained have already shown their importance with respect to both gender and age classification, it serves as a strong factor when compared to other cutting edge technologies.