Add dataset card, task category, and links to paper and code

README.md

@@ -1,3 +1,36 @@
----
-license: apache-2.0
----
+---
+license: apache-2.0
+task_categories:
+- image-to-text
+tags:
+- cad
+- 3d-reconstruction
+---
+
+# Ortho2CAD Orthographic Drawings
+
+This repository contains the orthographic drawings dataset from the paper [Ortho2CAD: 3D CAD generation from orthographic drawings using vision language models](https://huggingface.co/papers/2607.08891).
+
+- **GitHub Repository**: [AdityaJoglekar/Ortho2CAD](https://github.com/AdityaJoglekar/Ortho2CAD)
+- **Paper**: [Ortho2CAD: 3D CAD generation from orthographic drawings using vision language models](https://huggingface.co/papers/2607.08891)
+
+## Dataset Description
+
+Ortho2CAD is a dataset designed to facilitate research in translating rasterized orthographic drawings directly into editable CadQuery code, which can then be compiled into 3D CAD models.
+
+The dataset includes:
+- `ortho_train_data.zip`: Contains the training JSON file and the orthographic drawing images for the DeepCAD dataset.
+- `f360rec.zip`: Contains the training JSON file and the orthographic drawing images for the Fusion 360 Reconstruction dataset.
+
+For detailed setup, drawing generation, training, and evaluation scripts, please refer to the [official GitHub repository](https://github.com/AdityaJoglekar/Ortho2CAD).
+
+## Citation
+
+```bibtex
+@article{joglekar2026ortho2cad,
+  title={Ortho2CAD: 3D CAD generation from orthographic drawings using vision language models},
+  author={Joglekar, Aditya and Regmi, Amit and Shimada, Kenji and Kara, Levent Burak},
+  journal={arXiv preprint arXiv:2607.08891},
+  year={2026}
+}
+```