Convolutional Neural Network and Human Age Detection : A Pilot Study

The problem of age detection in age required job has become a challenging issue in contemporary times. Job seekers falsify their ages in order to secure jobs within the required age limits. Agetoolpie, one of the existing sophisticated age prediction systems available online, captures images improperly. Thus, this research is aimed at designing a deep model to predict age of an individual, using features from image. To achieve this aim, Cross Industrial Standard Process Data Mining (CRISP-DM) is the methodology adopted. Dataset acquired from Kaggle machine learning data repository was used to train & test the system, while the system predicts and from the captured image it predicts the age of individuals based on the Convolutional neural network (CNN or ConvNet) algorithm implemented on the system. The performance of the proposed system, ConvNet AgeDet, was compared with that of Agetoolpie in terms of precision, sensitivity, specificity, and accuracy. The result shows that the proposed system has a better age prediction accuracy of 71% than Agetoolpie’s 67%. The study concludes that artificial intelligence models are capable of solving issues of age falsification. The ConvNet AgeDet is recommended to organizations interested in getting the real ages of job applicants. Researchers are enjoined to delve more into AI-driven age detection and prediction.