Included in the blog posts, you will find two parts:
In Part I, I will provide you an overview of the Generative Adversarial Networks(GANs). In Part II, you will follow me to implement GANs techniques via an interesting synthetic image generation case. Tensorflow will be used as the backend.
To run the model:
Open the Jupyter Notebook from the command line, and navigate to ’Project_1_Fonts_Generation_using_GAN.ipynb’ and open it in the browser. You can run the code directly within the Jupyter/Ipython Notebook. Codes for downloading the dataset from the online source has been included in the notebook. However, in case the link gets invalid, the dataset to be used for the model has been included in a folder named ‘notMNIST_small’ within the data folder. Please check it out if the dataset cannot be downloaded from the internect by the codes automatically.
Feel free to test the codes and play around with GANs. Enjoy!