Improve dataset card: Add metadata, links, and sample usage

#2
by nielsr HF Staff - opened
Files changed (1) hide show
  1. README.md +12 -2
README.md CHANGED
@@ -1,6 +1,15 @@
 
 
 
 
 
 
1
  # Pixie Dataset
2
 
3
- This dataset contains data and pre-trained models for the Pixie project.
 
 
 
4
 
5
  ## Contents
6
 
@@ -9,7 +18,7 @@ This dataset contains data and pre-trained models for the Pixie project.
9
  - `real_scene_data/`: Real scene data for evaluation
10
  - `real_scene_models/`: Trained models for real scenes
11
 
12
- ## Usage
13
 
14
  Use the download script in the Pixie repository to automatically download this data:
15
 
@@ -17,3 +26,4 @@ Use the download script in the Pixie repository to automatically download this d
17
  python scripts/download_data.py
18
  ```
19
 
 
 
1
+ ---
2
+ task_categories:
3
+ - image-to-3d
4
+ license: mit
5
+ ---
6
+
7
  # Pixie Dataset
8
 
9
+ This dataset contains data and pre-trained models for the paper [Pixie: Fast and Generalizable Supervised Learning of 3D Physics from Pixels](https://huggingface.co/papers/2508.17437).
10
+
11
+ - Project Page: https://pixie-3d.github.io/
12
+ - Code: https://github.com/vlongle/pixie
13
 
14
  ## Contents
15
 
 
18
  - `real_scene_data/`: Real scene data for evaluation
19
  - `real_scene_models/`: Trained models for real scenes
20
 
21
+ ## Sample Usage
22
 
23
  Use the download script in the Pixie repository to automatically download this data:
24
 
 
26
  python scripts/download_data.py
27
  ```
28
 
29
+ For a full pipeline usage, refer to the [Github repository's usage section](https://github.com/vlongle/pixie#usage).