Update model card and pipeline tag

This PR improves the model card for Think2Seg-RS-3B. Key changes include:
- Updating the `pipeline_tag` to `image-segmentation` to better reflect the reasoning segmentation task.
- Adding a link to the [research paper](https://huggingface.co/papers/2512.19302).
- Linking the official [GitHub repository](https://github.com/Ricardo-XZ/Think2Seg-RS).
- Adding a BibTeX citation for the paper.

README.md

@@ -1,14 +1,42 @@
 ---
 base_model:
 - Qwen/Qwen2.5-VL-3B-Instruct
-pipeline_tag: image-text-to-text
 library_name: transformers
+pipeline_tag: image-segmentation
 ---
 
-## Bridging Semantics and Geometry: A Decoupled LVLM–SAM Framework for Reasoning Segmentation in Remote Sensing
+# Think2Seg-RS-3B
 
-This is the 3B model of [Think2Seg-RS](https://github.com/Ricardo-XZ/Think2Seg-RS), a decoupled framework for reasoning segmentation in remote sensing (RS) imagery.
+This repository contains the 3B prompter model for **Think2Seg-RS**, a decoupled framework for reasoning segmentation in remote sensing (RS) imagery.
 
-Our core idea is to decouple high-level semantic reasoning from low-level geometric execution. Specifically, we train an LVLM prompter (e.g., Qwen-2.5-VL) to control a frozen Segment Anything Model (SAM2) via structured geometric prompts. Through a result-oriented reinforcement learning objective, the LVLM learns to translate abstract semantic reasoning into spatially grounded actions, achieving state-of-the-art performance on the EarthReason dataset.
+## Overview
 
-For more details, code, and the complete framework, please visit our [GitHub repository](https://github.com/Ricardo-XZ/Think2Seg-RS).
+Think2Seg-RS addresses the limitations of coupling linguistic reasoning and pixel prediction in remote sensing analysis. The framework decouples high-level semantic reasoning from low-level geometric execution by training an LVLM prompter (based on Qwen2.5-VL) to control a frozen Segment Anything Model (SAM2) via structured geometric prompts. 
+
+Through a result-oriented reinforcement learning objective, the model learns to translate abstract semantic reasoning into spatially grounded actions, achieving state-of-the-art performance on the EarthReason dataset.
+
+- **Paper:** [Bridging Semantics and Geometry: A Decoupled LVLM-SAM Framework for Reasoning Segmentation in Remote Sensing](https://huggingface.co/papers/2512.19302)
+- **Repository:** [GitHub - Ricardo-XZ/Think2Seg-RS](https://github.com/Ricardo-XZ/Think2Seg-RS)
+- **Base Model:** [Qwen2.5-VL-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-3B-Instruct)
+
+## Key Features
+- **Decoupled Architecture:** Separates high-level semantic reasoning from low-level geometric execution.
+- **Geospatial Understanding:** Optimized for the complexities of remote sensing imagery and heterogeneous backgrounds.
+- **Zero-shot Generalization:** The learned prompting policy generalizes effectively across multiple referring segmentation benchmarks.
+
+## Setup and Usage
+
+For installation, training, and evaluation scripts, please visit the official [GitHub repository](https://github.com/Ricardo-XZ/Think2Seg-RS).
+
+## Citation
+
+If you find this work helpful, please consider citing:
+
+```bibtex
+@article{think2seg_rs_2025,
+  title={Bridging Semantics and Geometry: A Decoupled LVLM-SAM Framework for Reasoning Segmentation in Remote Sensing},
+  author={Luo, Junyu and Luo, Xiao and Chen, Xiusi and Xiao, Zhiping and Ju, Wei and Zhang, Ming},
+  journal={arXiv preprint arXiv:2512.19302},
+  year={2025}
+}
+```