Tahmini - Age and Gender Prediction Using Convolutional Neural Networks<bold> </bold>

Safak E., BARIŞÇI N.

2nd International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT), Kizilcahamam, Türkiye, 19 - 21 Ekim 2018, ss.348-354 identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası:
  • Basıldığı Şehir: Kizilcahamam
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.348-354
  • Gazi Üniversitesi Adresli: Evet

Özet

Age has always been an important feature of our identity. It is also an important factor in our social life. Predictions of age made with artificial intelligence can be applied to many areas such as intelligent human-machine interface development, security, cosmetics, electronic commerce. In this study, face images of persons were trained using convolutional neural networks, and age and gender were tried to be predicted with high success rate. Inception V1 convolutional neural network model developed by Google was used for training. Inception V1 model is already trained in the VGGFace2 dataset. Transfer learning provided by Deep Learning, the dataset is trained on this model. The IMDB dataset with face images with gender and age tags was selected as the dataset. There are 460.723 images in this dataset. 260.428 pictures in the IMDB dataset are used for training. As a result of the study, achievement rate of 70.3% in age prediction and 97% in gender prediction has been achieved.