Modified version of Taehoon Kim’s tensorflow implementation of DCGAN https://carpedm20.github.io/faces/ with a focus on generating paintings and soon graphic design.
It includes a script to scrape WikiArt, one to uniform images in a format that the Dcgan can work with and a Google Colab Notebook to train it on a free GPU.
- Python 2.7 or Python 3.3+
- Tensorflow 0.12.1
- SciPy
- pillow
You can find a zip of my dataset in: Dataset
Checkpoints here : Checkpoints
Colab Notebook: Colab Notebook:
You will have to convert the images to RGB and resize them with: uniform.py by changing path with your directory
Train your own dataset:
$ mkdir data/DATASET_NAME
... add images to data/DATASET_NAME ...
$ python main.py --dataset DATASET_NAME --train
$ python main.py --dataset DATASET_NAME
$ # example
$ python main.py --dataset=eyes --input_fname_pattern="*_cropped.png" --train
If your dataset is located in a different root directory:
$ python main.py --dataset DATASET_NAME --data_dir DATASET_ROOT_DIR --train
$ python main.py --dataset DATASET_NAME --data_dir DATASET_ROOT_DIR
$ # example
$ python main.py --dataset=eyes --data_dir ../datasets/ --input_fname_pattern="*_cropped.png" --train