download-celebA-HQ

Python script to download and create the celebA-HQ dataset.

WARNING from the author. I believe this script is broken since a few months (I have not try it for a while). I am really sorry about that. If you fix it, please share you solution in a PR so that everyone can benefit from it.

To get the celebA-HQ dataset, you need to a) download the celebA dataset download_celebA.py , b) download some extra files download_celebA_HQ.py, c) do some processing to get the HQ images make_HQ_images.py.

The size of the final dataset is 89G. However, you will need a bit more storage to be able to run the scripts.

Usage

1) Clone the repository

git clone https://github.com/nperraud/download-celebA-HQ.git
cd download-celebA-HQ

2) Install necessary packages (Because specific versions are required Conda is recomended)

• Install miniconda https://conda.io/miniconda.html
• Create a new environement

  conda create -n celebaHQ python=3
  source activate celebaHQ

• Install the packages

  conda install jpeg=8d tqdm requests pillow==3.1.1 urllib3 numpy cryptography scipy
  pip install opencv-python==3.4.0.12 cryptography==2.1.4

• Install 7zip (On Ubuntu)

  sudo apt-get install p7zip-full

3) Run the scripts

python download_celebA.py ./
python download_celebA_HQ.py ./
python make_HQ_images.py ./

where ./ is the directory where you wish the data to be saved.

4) Go watch a movie, theses scripts will take a few hours to run depending on your internet connection and your CPU power. The final HQ images will be saved as .npy files in the ./celebA-HQ folder.

Windows

The script may work on windows, though I have not tested this solution personnaly

Step 2 becomes

• Install miniconda https://conda.io/miniconda.html or anaconda
• Create a new environement

  conda create -n celebaHQ python=3
  source activate celebaHQ

• Install the packages

  conda  install -c anaconda jpeg=8d tqdm requests pillow==3.1.1 urllib3 numpy cryptography scipy

• Install 7zip

The rest should be unchanged.

Docker

If you have Docker installed, skip the previous installation steps and run the following command from the root directory of this project:

docker build -t celeba . && docker run -it -v $(pwd):/data celeba

By default, this will create the dataset in same directory. To put it elsewhere, replace $(pwd) with the absolute path to the desired output directory.

Outliers

It seems that the dataset has a few outliers. A of problematic images is stored in bad_images.txt. Please report if you find other outliers.

Remark

This script is likely to break somewhere, but if it executes until the end, you should obtain the correct dataset.

Sources

This code is inspired by these files

• https://github.com/tkarras/progressive_growing_of_gans/blob/master/dataset_tool.py
• https://github.com/andersbll/deeppy/blob/master/deeppy/dataset/celeba.py
• https://github.com/andersbll/deeppy/blob/master/deeppy/dataset/util.py

Citing the dataset

You probably want to cite the paper "Progressive Growing of GANs for Improved Quality, Stability, and Variation" that was submitted to ICLR 2018 by Tero Karras (NVIDIA), Timo Aila (NVIDIA), Samuli Laine (NVIDIA), Jaakko Lehtinen (NVIDIA and Aalto University).