Dataset · a1557811266
Inter-Edit image editing dataset
a1557811266/Inter-Edit-Train
Official dataset for interactive image editing based on text instructions and spatial cues.
What's inside
Contains 1,099,964 image editing pairs and 610,186 unique source images in English and Chinese. Total dataset size falls between 1M and 10M records. Designed for the Interactive Instruction-based Image Editing (I^3E) task from a CVPR 2026 paper.
What you can do with it
- Training local text-guided image editing models
- Developing interactive computer vision systems
- Tasks
- Image → image
- Languages
- English, Chinese
- Size
- 1M–10M rows
- Downloads this month
- 20K
- Likes
- 2
How to load it
from datasets import load_dataset
ds = load_dataset("a1557811266/Inter-Edit-Train")