Update README.md

README.md

@@ -1,3 +1,43 @@
+
 ---
 license: cc-by-nc-4.0
 ---
+# SpatialForge: Bootstrapping 3D-Aware Spatial Reasoning from Open-World 2D Images
+
+🤗 [Hugging Face](https://huggingface.co/datasets/shana643/SpatialForge) | 
+📑 [Paper](https://arxiv.org/html/2605.11462v1) | 
+
+Zishan Liu, Ruoxi Zang, Yanglin Zhang, Wei Liu, Yin Zhang, Jian Yao, Jiayin Zheng, Zhengzhe Liu
+
+*Lingnan University · XPENG Robotics*
+
+The official data release for **SpatialForge-10M**: A Large-Scale Spatial Reasoning Dataset Bootstrapped from Open-World 2D Images.
+
+---
+
+## 📌 Important Notice
+
+This is the **full release** of SpatialForge-10M, containing **10 million** spatial QA pairs derived from **2.8 million** curated open-world images across six spatial tasks.
+
+**Key Features:**
+- ✅ **10M spatial QA pairs** spanning perception (grounding, referring, counting) and relation (near-far, left-right, perspective)
+- ✅ **Open-world diversity** — 2.8M images from Objects365, OpenImages, Pixmo
+- ✅ **CC BY-NC 4.0 license** — free for academic and non-commercial use
+
+> **Note**: All bounding boxes are normalized to `[0, 1000]` to align with Qwen3-VL's pre-training format.
+
+---
+
+## 🎯 Task Overview
+
+| Level | Task | Description | Count |
+|-------|------|-------------|-------|
+| **Perception** | Grounding | Localize objects from descriptions → bbox | 3.6M |
+| | Referring | Describe objects within a bbox | 3.6M |
+| | Counting | Count objects satisfying a condition | 495k |
+| **Relation** | Near-Far | Determine relative depth ordering | 2.6M |
+| | Left-Right | Identify horizontal relations (camera-centric) | 93k |
+| | Perspective | Reason from human-centric viewpoints | 8k |
+| **Total** | | | **10.2M** |
+
+---