EVALUATION OF SIMULTANEOUS IDENTITY, AGE AND GENDER RECOGNITION FOR CROWD FACE MONITORING

Nowadays, facial recognition combined with age estimation and gender prediction has been deeply involved the factors associated crowd monitoring. This is considered to be a major complex job for humans. paper proposes unified system based on already available deep learning machine models (i.e., FaceNet, ResNet, Support Vector Machine, AgeNet GenderNet) that automatically simultaneously performs person identification, prediction. Then evaluated newly proposed multi-face, realistic challenging test dataset. The current face technology primarily focuses static datasets of known identities does not focus novel identities. approach suitable continuous In our system, whenever are found during inference, will save those an appropriate label each unique identity updated periodically in order correctly recognise future inference iterations. However, extracting features whole dataset new detected efficient solution. To address this issue, we propose incremental feature extraction training method which aims reduce computational load extraction. When tested dataset, recognizes pre-trained identities, estimates age, predicts average accuracy 49%, 66.5% 93.54% respectively. We conclude can sensitive robust uncontrolled environment (e.g., abrupt lighting conditions).