Congitional-GAN

Conditional GAN trained to create a car images of size 36x36

(scientific paper about cGANs is in file Paper.pdf)

• Input dataset is in "Input_car_train" path.
• All the code is contained in the Jupyter-notebook file "cGAN_cars.ipynb"
• File "cGAN_cars.py" is the Jupyter file converted to *.py format
• File "cGAN_cars.html" is the Jupyter file converted to *.html format to read it as a simple page in your browser
• Path "tensorboards" contains all scalars, graphs and histograms presented in "Report_Orel_1836231.pdf"
• Path "out" contains intermediate training results at every 1000 training steps
• Path "model_weights" contains all the model weights for the trained model. This path is not presented here in GitHub, you can find it in my cloud storage: https://yadi.sk/d/-rQfGuyK_gYzEQ