Update model card metadata and pipeline tag

Hi! I'm Niels from the Hugging Face community science team. This PR updates the model card metadata by:
- Adding the `pipeline_tag: image-to-3d` to help users find your model in the correct category.
- Removing the `arxiv` field from the YAML header to follow the Hub's best practices (keeping the reference in the Markdown section).

README.md

@@ -1,23 +1,23 @@
 ---
-license: apache-2.0
 language:
-  - en
+- en
+license: apache-2.0
+pipeline_tag: image-to-3d
 tags:
-  - 3d-reconstruction
-  - depth-estimation
-  - camera-pose-estimation
-  - streaming
-  - video-understanding
-  - computer-vision
-  - point-cloud
-arxiv: 2605.23889
+- 3d-reconstruction
+- depth-estimation
+- camera-pose-estimation
+- streaming
+- video-understanding
+- computer-vision
+- point-cloud
 ---
 
 # HorizonStream
 
 **HorizonStream: Long-Horizon Attention for Streaming 3D Reconstruction**
 
-[[Paper](https://arxiv.org/abs/2605.23889)] [[Code](https://github.com/3dagentworld/horizonstream)] [[Project Page](https://3dagentworld.github.io/horizonstream/)]
+[[Paper](https://arxiv.org/abs/2605.23889)] [[Code](https://github.com/3DAgentWorld/HorizonStream)] [[Project Page](https://3dagentworld.github.io/horizonstream/)]
 
 ## Overview
 
@@ -54,5 +54,4 @@ python infer.py \
       primaryClass={cs.CV},
       url={https://arxiv.org/abs/2605.23889}, 
 }
-```
-
+```