Update metadata (language, license, task categories) and add code/project links

This PR updates the dataset card with accurate metadata and supplementary links:

- The `language` tag in the metadata is corrected from `aa` to `en`.
- The `license` in the metadata is updated from `apache-2.0` to `mit`, aligning with the license specified in the associated GitHub repository.
- `task_categories: ['image-to-3d', 'image-segmentation']` are added to improve discoverability, reflecting the dataset's use in semantic 3D reconstruction and instance-level mask annotation tasks.
- Links to the official GitHub repository (`https://github.com/lifuguan/IGGT_official`) and the project page (`https://lifuguan.github.io/IGGT_official`) are integrated into the content for easy access to supplementary materials. The existing arXiv paper link remains unchanged.

README.md

@@ -1,7 +1,10 @@
 ---
-license: apache-2.0
 language:
-- aa
+- en
+license: mit
+task_categories:
+- image-to-3d
+- image-segmentation
 ---
 
 This is the official dataset `InsScene-15K` for the paper:
@@ -11,9 +14,9 @@ This is the official dataset `InsScene-15K` for the paper:
 >
 > <sup>*</sup>Equal Contribution, <sup>†</sup>Project Leader, Corresponding author.
 >
-> ### [Paper](https://arxiv.org/abs/2510.22706)
+> ### [Paper](https://arxiv.org/abs/2510.22706) | [Project Page](https://lifuguan.github.io/IGGT_official) | [Code](https://github.com/lifuguan/IGGT_official)
 
 
 ![iggt_teaser](https://cdn-uploads.huggingface.co/production/uploads/667b8de7a68bf81afe668afe/KiYMUhU08esa0eGAoRRF0.png)
 
-## Please note the datasets are still being uploaded.
+## Please note the datasets are still being uploaded.