Datasets:

---

smz8599

/

UGround-single

like 0

This PR adds an initial, comprehensive dataset card for the GUI-AIMA dataset.

It includes:

• Relevant metadata such as task_categories: ['image-text-to-text'], language: ['en'], and tags like gui-grounding and multimodal for improved discoverability.
• A direct link to the associated paper on Hugging Face: GUI-AIMA: Aligning Intrinsic Multimodal Attention with a Context Anchor for GUI Grounding.
• Links to the project page and GitHub repository.
• The full abstract of the paper to provide detailed context.
• A section on data preparation, outlining the external datasets required for training models with this data.
• A "Sample Usage" section featuring installation instructions directly from the GitHub README and referencing an inference script.
• The BibTeX citation for proper academic referencing.