Human Age Prediction Based on Hand Image using Multiclass Classification

The human identification/recognition based on biometric data (face, voice, fingerprint, etc.) is the subject of various studies. It can be improved by introducing the human age factor. This paper proposes a method of human age prediction based on hand image by using a multiclass classification. The architecture used is a time distributed convolutional neural network combined with a Gated Recurrent Unite followed by a stack of fully connected layers, which extract hand skin features from a different aspect of hand Dorsal Left, Palmar Left, Dorsal Right, and Palmar Right. The results are a classification into 17 classes from range age 18 - 75. Experimental results show that the proposed method has effective performance in age prediction from hand images with an accuracy of 96.5%.