Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,25 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: mit
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
---
|
| 4 |
+
|
| 5 |
+
# **Fine-Grained Preference Optimization Improves Spatial Reasoning in VLMs**
|
| 6 |
+
|
| 7 |
+
______________________________________________________________________
|
| 8 |
+
|
| 9 |
+
## 💡 Introduction
|
| 10 |
+
<div align="center">
|
| 11 |
+
|
| 12 |
+
<a href="https://plan-lab.github.io/projects/spatialreasoner/"><img src="https://img.shields.io/badge/Project-Page-blue?style=for-the-badge&logo=googlechrome&logoColor=white"></a>
|
| 13 |
+
<a href="https://arxiv.org/pdf/2506.21656"><img src="https://img.shields.io/badge/Arxiv-2506.21656-red?style=for-the-badge&logo=arxiv&logoColor=white"></a>
|
| 14 |
+
<a href="https://huggingface.co/PLAN-Lab/SpatialReasoner-R1"><img src="https://img.shields.io/badge/Hugging%20Face-Model-yellow?style=for-the-badge&logo=huggingface&logoColor=black"></a>
|
| 15 |
+
|
| 16 |
+
</div>
|
| 17 |
+
|
| 18 |
+
**Yifan Shen, Yuanzhe Liu, Jingyuan Zhu, Xu Cao, Xiaofeng Zhang, Yixiao He, Wenming Ye, James Matthew Rehg, Ismini Lourentzou**
|
| 19 |
+
|
| 20 |
+
Current Vision-Language Models (VLMs) struggle with fine-grained spatial reasoning, particularly when multi-step logic and precise spatial alignment are required. In this work, we introduce SpatialReasoner-R1, a novel VLM designed to address these limitations. First, we propose Multi-LLM Guided Monte Carlo Tree Search (M3CTS) and Fine-Grained Spatial Rewards methods to construct a high-quality dataset. Second, we use fine-grained Direct Preference Optimization (fDPO) to train our model. fDPO introduces segment-specific preference granularity for descriptive grounding and logical reasoning, achieving an average improvement of 4.1% over standard DPO across spatial quality tasks, and a 9.0% boost in spatial quantity tasks. To address the scarcity of multi-step spatial reasoning data, M3CTS enables collaborative exploration of diverse reasoning paths, significantly enriching spatial comprehension and logical coherence. Empirical evaluations demonstrate that SpatialReasoner-R1 sets a new state-of-the-art on SpatialRGPT-Bench, outperforming the strongest baseline by 9.8% in average accuracy, while maintaining competitive performance on general vision-language tasks.
|
| 21 |
+
|
| 22 |
+
|
| 23 |
+
## Model Card Contact
|
| 24 |
+
|
| 25 |