Datasets:
Dataset Summary
This dataset was obtained from the original CycleGAN Datasets directory available on Berkeley's website.
For more details about the dataset you can refer to the original CycleGAN publication.
How to use
You can easily load the dataset with the following lines :
from datasets import load_dataset
data_horses = load_dataset("gigant/horse2zebra", name="horse", split="train")
data_zebras = load_dataset("gigant/horse2zebra", name="zebra", split="train")
Two splits are available, "train"
and "test"
Citation Information
@inproceedings{CycleGAN2017,
title={Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks},
author={Zhu, Jun-Yan and Park, Taesung and Isola, Phillip and Efros, Alexei A},
booktitle={Computer Vision (ICCV), 2017 IEEE International Conference on},
year={2017}
}
- Downloads last month
- 132
Homepage:
people.eecs.berkeley.edu
Paper:
arxiv.org
Size of downloaded dataset files:
234 MB
Size of the auto-converted Parquet files:
343 MB
Number of rows:
2,661