Romulan12 / Face-Generation-Using-DCGAN

Training a DCGAN on a dataset of faces to get the generator network to generate new images that look like realistic faces.

Face-Generation-Using-DCGAN

Training a DCGAN on a dataset of faces to get the generator network to generate new images that look like realistic faces. The model was trained on a subset of the celebA dataset.

Architecture

Usage

Train :

pyhton3 train.py

Test

pyhton3 test.py

Visualising Loss

Plot of the training losses for the generator and discriminator:

Results