Add dataset card and metadata for GUI-AIMA

README.md

@@ -0,0 +1,38 @@
+---
+license: other
+task_categories:
+- image-text-to-text
+tags:
+- gui-grounding
+- computer-use
+- multimodal
+---
+
+# GUI-AIMA Dataset
+
+This repository contains the training dataset for **GUI-AIMA: Aligning Intrinsic Multimodal Attention with a Context Anchor for GUI Grounding**. 
+
+- **Paper:** [GUI-AIMA: Aligning Intrinsic Multimodal Attention with a Context Anchor for GUI Grounding](https://huggingface.co/papers/2511.00810)
+- **Repository:** [https://github.com/sjz5202/GUI-AIMA](https://github.com/sjz5202/GUI-AIMA)
+
+## Dataset Description
+
+This dataset is compiled for training the GUI-AIMA models (specifically GUI-AIMA-lite) and consists of:
+1. **GUI-Actor data** from [cckevinn/GUI-Actor-Data](https://huggingface.co/datasets/cckevinn/GUI-Actor-Data)
+2. **UGround single-round dialogue data** from [smz8599/UGround-single](https://huggingface.co/datasets/smz8599/UGround-single)
+3. **GTA1 data** (excluding the web portion) from [smz8599/GTA_data_no_web](https://huggingface.co/datasets/smz8599/GTA_data_no_web)
+
+For GUI-AIMA-3B, the model was further trained with an additional 250k data from [GroundCUA](https://huggingface.co/datasets/ServiceNow/GroundCUA).
+
+## Citation
+
+If you use this dataset or the GUI-AIMA models, please cite the following paper:
+
+```bibtex
+@article{zhou2025gui,
+  title={Gui-aima: Aligning intrinsic multimodal attention with a context anchor for gui grounding},
+  author={Zhou, Shijie and Lai, Viet Dac and Tan, Hao and Kil, Jihyung and Zhu, Wanrong and Chen, Changyou and Zhang, Ruiyi},
+  journal={arXiv preprint arXiv:2511.00810},
+  year={2025}
+}
+```