Hand Image-Based Human Age Estimation using a Time Distributed CNN-GRU
Mohamed Ait Abderrahmane, Ibrahim Guelzim, Abdelkaher Ait Abdelouahad
20202020 International Conference on Data Analytics for Business and Industry: Way Towards a Sustainable Economy (ICDABI)12 citationsDOI
Abstract
The prediction of human age is more and more widely used in various fields such as forensic medicine or national security. Several studies based on biometrics (face, voice, fingerprint) have been conducted. This paper proposes an age estimation method based on the color image of the hand using a time distributed convolutional neural network combined with Gated Recurrent Unite (GRU) followed by a stack of fully connected layers. An adjustment is used to eliminate model bias caused by the gender and the skin color, experiments demonstrate good performance of the model.
Topics & Concepts
Convolutional neural networkComputer scienceBiometricsArtificial intelligenceFace (sociological concept)Fingerprint (computing)Image (mathematics)Computer visionDeep learningPattern recognition (psychology)Social scienceSociologyFace recognition and analysisBiometric Identification and SecurityGait Recognition and Analysis