Abstract:
Gender classification and Age identification play an important role in our
social lives. Gender is central characteristics of personality, and it is
essential in our life. Age is important for our identity. Security, biometric
system, and treatment are part of gender classification and age prediction.
Age prediction can help to authorize people from buying adult products or
other kind of restricted goods. In this project, for classification of gender
and prediction of age from pictures deep learning model is used. The
objective of this study is to create a model for gender classification and an
age estimation using convolutional neural networks and ResNet50. The
image's feature extraction and categorization are included by CNN. Feature
extraction gives the features corresponding to gender and age from the face
pictures whereas the classification classify the image into correct age and
gender.ResNet50 is the convolutional network that have 50 layers. Age
prediction is the regression problem and prediction of gender is a binary
classification problem. The model is evaluated using the UTKFace dataset,
a sizable face dataset with a broad age range. Deep learning algorithm is
used to obtain higher accuracy and lower MAE, also MAE of the both
algorithm is compared to obtain which algorithm more efficient.
