Add image-to-video task category and Github link

This PR updates the `task_categories` to reflect that this dataset is used for image-to-video generation. It also adds a link to the GitHub repository for easier access to the code.

README.md

@@ -2,15 +2,17 @@
 license: mit
 tags:
 - renderformer
+task_categories:
+- image-to-video
 ---
 
 # RenderFormer Video Data
 
-This repo contains the example data for RenderFormer video rendering.
+This repo contains the example data for RenderFormer video rendering.  The data includes various 3D models and scenes for testing.
 
 ## Quick Start
 
-Please refer to our [Github Repo](https://github.com/microsoft/renderformer) and [Paper](https://arxiv.org/abs/2505.21925).
+Please refer to our [Github Repo](https://github.com/microsoft/renderformer) and [Paper](https://huggingface.co/papers/2505.21925).
 
 ## Acknowledgement
 
@@ -45,4 +47,4 @@ If you find our repository useful, please cite our paper in your work:
     booktitle  = {ACM SIGGRAPH 2025 Conference Papers},
     year       = {2025}
 }
-```
+```