Keras Text to Image Generation Project

Text to Image using GloVe and Word2Vec with Generative Adversarial Network Models in Keras.

Thanks to : Xianshun Chen where his original project generating Pokémon images which provided the bases for this project.

Xianshun's original Project here - https://github.com/chen0040/keras-text-to-image

Note : I have streamlined the original project and picked one method of generating the model with keras while I continue to unpick how the model is being generated.

Usage

• Training Image files should be placed in /data/train_images
• Training Text files should be placed in /data/train_text
• Execute dcgan_v3_train_BathML.py (Highly recomend using a GPU tensorflow install detailed below)
• Text to be used for generation should be placed in the location /data/generate_text (Separate text files will generate separate images)
• Execute dcgan_v3_generate_BathML.py (This doesn't require a GPU tensorflow install and I tend to use one machine for Training and a lighter machine for generation where the model files should be transferred) & generated images should be written to the location /data/generated_images

Configure to run on GPU on Windows

• Step 1: Change tensorflow to tensorflow-gpu in requirements.txt and install tensorflow-gpu
• Step 2: Download and install the CUDA® Toolkit 9.0 (Please note that currently CUDA® Toolkit 9.1 is not yet supported by tensorflow, therefore you should download CUDA® Toolkit 9.0)
• Step 3: Download and unzip the cuDNN 7.4 for CUDA@ Toolkit 9.0 and add the bin folder of the unzipped directory to the $PATH of your Windows environment