Update README.md

README.md

@@ -1,8 +1,18 @@
+---
+license: apache-2.0
+datasets:
+- usm3d/hoho25k
+language:
+- en
+tags:
+- hoho25k
+- s23dr2025
+---
 # S23DR 2025 Challenge - Winning Solution 🏆
 
 This repository contains the **winning solution** for the [Structured Semantic 3D Reconstruction (S23DR) Challenge 2025](https://huggingface.co/spaces/usm3d/S23DR2025) at CVPR 2025 Workshop.
 
-Our method achieves state-of-the-art performance in 3D wireframe reconstruction from multi-view images, combining COLMAP point clouds, semantic segmentation, and deep learning models to predict building wireframes with high accuracy.
+Our method achieves the winning performance in 3D wireframe reconstruction from multi-view images, combining COLMAP point clouds, semantic segmentation, and deep learning models to predict building wireframes with high accuracy.
 
 ## 🎯 Performance
 
@@ -160,12 +170,7 @@ We evaluate using three key metrics:
 Please cite our work if you use this code:
 
 ```bibtex
-@inproceedings{skvrna2025s23dr,
-  title={Winning Solution for S23DR Challenge 2025: Multi-Modal 3D Wireframe Reconstruction},
-  author={Škvrňa, Jan},
-  booktitle={CVPR 2025 Workshop on Structured Semantic 3D Reconstruction},
-  year={2025}
-}
+TODO
 ```
 
 ## 📋 Requirements
@@ -190,4 +195,4 @@ Thanks to the S23DR Challenge organizers and the HoHo25k dataset creators for pr
 
 ---
 
-For detailed technical description, please refer to our paper and the comprehensive code documentation throughout the repository.
+For detailed technical description, please refer to our paper and the comprehensive code documentation throughout the repository.