Flickr-Faces-HQ (FFHQ) downsampled to 32x32 resolution.
Original full-resolution dataset is taken from [
https://github.com/NVlabs/ffhq-dataset].
@inproceedings{karras2019style,
title={A style-based generator architecture for generative adversarial networks},
author={Karras, Tero and Laine, Samuli and Aila, Timo},
booktitle={Proceedings of the IEEE/CVF conference on computer vision and pattern recognition},
pages={4401--4410},
year={2019}
}
