Spaces:
Running on Zero
Running on Zero
| title: GPD Spatial Reasoning | |
| emoji: 🧭 | |
| colorFrom: pink | |
| colorTo: indigo | |
| sdk: gradio | |
| sdk_version: 6.30.0 | |
| app_file: app.py | |
| short_description: Spatial-reasoning VLM distilled with routed 3D privilege | |
| python_version: "3.12" | |
| # 🧭 GPD · Distilling Routed 3D Privilege for Spatial Reasoning | |
| This Space runs **GPD-4B**, an RGB-only vision-language model fine-tuned from | |
| [Qwen3-VL-4B-Instruct](https://huggingface.co/Qwen/Qwen3-VL-4B-Instruct) with | |
| **GPD** (*Geometry-Privileged Distillation*): on-policy self-distillation in which a | |
| teacher version of the model sees question-routed textual 3D cues (depth / semantic / | |
| bird's-eye-view) plus the reference answer, while the deployed student stays **RGB-only**. | |
| The privileged KL on incorrect trajectories augments GRPO. | |
| Upload a single image, ask a 3D spatial-reasoning question — observer-shift spatial | |
| imagination, directional relations between objects, depth / distance estimation in meters — | |
| and the model thinks step by step, then wraps its final answer in `<answer>` tags. | |
| ## Model | |
| - **Checkpoint:** [xinyili0624/GPD-4B](https://huggingface.co/xinyili0624/GPD-4B) (Apache-2.0) | |
| - **2B variant:** [xinyili0624/GPD-2B](https://huggingface.co/xinyili0624/GPD-2B) | |
| - **Paper:** [Distilling Routed 3D Privilege for Spatial Reasoning in Vision-Language Models](https://huggingface.co/papers/2610.12355) | |
| - **Code:** [github.com/ZJU-REAL/GPD](https://github.com/ZJU-REAL/GPD) | |
| - **Training data:** [xinyili0624/GPD-15k](https://huggingface.co/datasets/xinyili0624/GPD-15k) (VSI 10k + SPAR 4k + MindCube 1k) | |
| On Qwen3-VL-4B, GPD reaches **57.1** on VSI-Bench and **37.6** average across | |
| MindCube, SPAR-Bench, MMSI-Bench, and ViewSpatial. | |
| ## Examples | |
| The bundled example images and questions are taken from the authors' | |
| [GPD-15k](https://huggingface.co/datasets/xinyili0624/GPD-15k) dataset | |
| (Apache-2.0), specifically its SPAR single-image scenes, which are rendered with | |
| bounding-box / point annotations marking the objects the question refers to. | |
| ## Citation | |
| ```bibtex | |
| @article{li2026gpd, | |
| title={Distilling Routed 3D Privilege for Spatial Reasoning in Vision-Language Models}, | |
| author={Li, Hongxing and Li, Yixin and Li, Dingming and Wang, Zixuan and Yan, Yuchen and Zhang, Wenqi and Lu, Weiming and Shen, Yongliang}, | |
| journal={arXiv preprint arXiv:2610.12355}, | |
| year={2026} | |
| } | |
| ``` |