It was aimed to determine the beard on the human face and to replace the beard with beardless skin. I used the facial landmarks function of the dlib library to detect the chin on the face. I trained a classifier model to detect whether there is a beard on the chin image I took. I trained a GAN model to replace bearded images with appropriate beardless human skin. While training this model, I used the Plural Image Completion project developed by Chuanxia Zheng, Tat-Jen Cham and Jianfei Cai.
For more detailed information: https://github.com/CagriYonca/finalProject/blob/master/final_project.pdf
