Dataset · KORMo-VL
Ko-Screenspot Korean GUI grounding benchmark
KORMo-VL/Ko-Screenspot
A dataset of 3,319 screenshot-instruction pairs for GUI grounding tasks on Korean websites.
What's inside
Contains 3,319 samples (screenshot-instruction pairs) across 142 websites, derived from 540 human-recorded tasks. Instructions are written in Korean, and each sample includes target element annotations.
What you can do with it
- Training and evaluation of visual positioning (GUI grounding) models
- Developing multimodal assistants for Korean interfaces
- Tasks
- Image text → text
- Languages
- Korean
- Size
- up to 1K rows
- Downloads this month
- 174
- Likes
- 59
How to load it
from datasets import load_dataset
ds = load_dataset("KORMo-VL/Ko-Screenspot")