Dataset · ImagenHub
Text-guided image editing benchmark
ImagenHub/Text_Guided_Image_Editing
Set of text-image pairs from the ImagenHub project for standardized evaluation of image editing models.
What's inside
The dataset contains less than 1,000 examples with images and English text descriptions. It was created by the authors of the ImagenHub paper for unified quality verification of conditional generation and image-to-image editing models.
What you can do with it
- Evaluating and comparing image editing models using text instructions
- Validating custom image-to-image pipelines against a single standard
- Tasks
- Image → image
- Languages
- English
- Size
- up to 1K rows
- Downloads this month
- 361
- Likes
- 34
How to load it
from datasets import load_dataset
ds = load_dataset("ImagenHub/Text_Guided_Image_Editing")