A truncated deep neural network for identifying age groups in real time images. Issue 3 (3rd April 2022)

• Record Type: Journal Article
• Authors: Gupta, Vedika; Dass, Pranav; Bansal, Vibhuti; Arora, Rameshwar
• Is Part Of: Journal of interdisciplinary mathematics. Volume 25:Issue 3(2022)
• Page Start: 851
• Page End: 861
• Publication Date: 2022-04-03
• Subjects: 92B20 -- 68T05; Mask regional convolutional neural network (Mask RCNN) -- Deep convolutional neural network (DCNN) -- Residual network (ResNet); Mathematics -- Periodicals; Mathematics; Periodicals; 510.5
• DOI: 10.1080/09720502.2021.2016917
• Languages: English
• ISSNs: 0972-0502

Abstract

Abstract: Recent research works have been focussing on estimating age from facial images. Age estimation from faces basically involves two sub-processes: extracting features and estimating learning function. Age classification from an input image is the task at hand in this project; age will be classified into 3 categories: 1) Toddler, 2) Teen, 3) Adult. Classifying age automatically from an image has been widely used in our day-to-day lives, particularly in the listed fields: biometrics, surveillance systems, and commercial kiosks. The purpose of this study is to categorize facial images based on their age. Prominently, previously existing research works were performed on contrived and unreal images curated in laboratories. Those images did not correctly portray the distinctions and fluctuations that are evident in real human faces. This paper uses deep convolutional neural networks (CNN) on the available data to overcome the above discussed challenge.