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

#1
by nielsr HF Staff - opened
Files changed (1) hide show
  1. README.md +65 -11
README.md CHANGED
@@ -1,9 +1,22 @@
1
  ---
2
  license: cc-by-nc-4.0
 
 
 
 
 
 
 
 
 
 
 
3
  ---
4
 
5
  # SAGE-3D InteriorGS USDZ: USDZ-Format 3D Gaussian Scenes for Isaac Sim
6
 
 
 
7
  InteriorGS dataset converted to USDZ format for seamless integration with NVIDIA Omniverse and Isaac Sim platforms.
8
 
9
  <div align="center">
@@ -55,27 +68,68 @@ InteriorGS_usdz/
55
  └── ... # 1,000 scenes total
56
  ```
57
 
 
58
 
 
59
 
60
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
61
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
62
 
 
63
 
64
  ## πŸ”— Related Datasets
65
 
66
  This dataset is part of the **SAGE-3D** project:
67
 
68
- 1. **InteriorGS**: Original 3DGS scenes with semantic annotations
69
- β†’ [spatialverse/InteriorGS](https://huggingface.co/datasets/spatialverse/InteriorGS)
70
 
71
- 2. **SAGE-3D InteriorGS USDZ** (This dataset): USDZ format for Isaac Sim
72
- β†’ [spatialverse/SAGE-3D_InteriorGS_usdz](https://huggingface.co/datasets/spatialverse/SAGE-3D_InteriorGS_usdz)
73
 
74
- 3. **SAGE-3D Collision Mesh**: Physics-enabled collision bodies
75
- β†’ [spatialverse/SAGE-3D_Collision_Mesh](https://huggingface.co/datasets/spatialverse/SAGE-3D_Collision_Mesh)
76
 
77
- 4. **SAGE-3D VLN Data**: Navigation trajectories and instructions
78
- β†’ [spatialverse/SAGE-3D_VLN_Data](https://huggingface.co/datasets/spatialverse/SAGE-3D_VLN_Data)
79
 
80
  ---
81
 
@@ -99,7 +153,7 @@ If you use SAGE-3D InteriorGS USDZ in your research, please cite:
99
  ```bibtex
100
  @misc{miao2025physicallyexecutable3dgaussian,
101
  title={Towards Physically Executable 3D Gaussian for Embodied Navigation},
102
- author={Bingchen Miao and Rong Wei and Zhiqi Ge and Xiaoquan Sun and Shiqi Gao and Jingzhe Zhu and Renhan Wang and Siliang Tang and Jun Xiao and Rui Tang and Juncheng Li},
103
  year={2025},
104
  eprint={2510.21307},
105
  archivePrefix={arXiv},
@@ -123,4 +177,4 @@ Please also cite the InteriorGS dataset:
123
 
124
  <div align="center">
125
  <strong>SAGE-3D: Semantically and Physically-Aligned Gaussian Environments for 3D Navigation</strong>
126
- </div>
 
1
  ---
2
  license: cc-by-nc-4.0
3
+ task_categories:
4
+ - robotics
5
+ - image-to-3d
6
+ language: en
7
+ tags:
8
+ - 3d-gaussian-splatting
9
+ - vln
10
+ - embodied-navigation
11
+ - isaac-sim
12
+ - omniverse
13
+ - usdz
14
  ---
15
 
16
  # SAGE-3D InteriorGS USDZ: USDZ-Format 3D Gaussian Scenes for Isaac Sim
17
 
18
+ [Paper](https://huggingface.co/papers/2510.21307) | [Project Page](https://sage-3d.github.io/) | [Code](https://github.com/Galery23/SAGE-3D_Official)
19
+
20
  InteriorGS dataset converted to USDZ format for seamless integration with NVIDIA Omniverse and Isaac Sim platforms.
21
 
22
  <div align="center">
 
68
  └── ... # 1,000 scenes total
69
  ```
70
 
71
+ ---
72
 
73
+ ## ✨ Sample Usage
74
 
75
+ This section demonstrates how to use this USDZ dataset within the SAGE-3D framework, specifically how to build USDA scene files with integrated collision bodies for Isaac Sim.
76
+
77
+ **Prerequisites:**
78
+ 1. Install `splat-transform` and `3DGRUT` as described in the [project's GitHub repository](https://github.com/Galery23/SAGE-3D_Official#prerequisites).
79
+ 2. Download the original [InteriorGS dataset](https://huggingface.co/datasets/spatialverse/InteriorGS) (compressed PLY files) and the [SAGE-3D Collision Mesh Dataset](https://huggingface.co/datasets/spatialverse/SAGE-3D_Collision_Mesh).
80
+
81
+ **Step 1: Convert Compressed PLY to Original PLY (if starting from original InteriorGS)**
82
+
83
+ ```bash
84
+ # Example: Convert a single scene
85
+ splat-transform /path/to/InteriorGS/0001_839920/3dgs_compressed.ply \
86
+ /path/to/output/0001_839920.ply
87
+ ```
88
+
89
+ **Step 2: Convert PLY to USDZ (if starting from original InteriorGS)**
90
 
91
+ ```bash
92
+ # Example: Convert a single scene
93
+ python -m threedgrut.export.scripts.ply_to_usd \
94
+ /path/to/ply/0001_839920.ply \
95
+ --output_file /path/to/usdz/0001_839920.usdz
96
+ ```
97
+ *Note: This dataset (`spatialverse/SAGE-3D_InteriorGS_usdz`) already provides the USDZ files, so you can directly use them instead of performing Steps 1 and 2.*
98
+
99
+ **Step 3: Build USDA Scene Files with Collision Meshes**
100
+
101
+ To use the USDZ scenes with physics simulation in Isaac Sim, they need to be combined with collision meshes and converted to USDA format.
102
+
103
+ ```bash
104
+ python Code/benchmark/scene_data/sage3d_usda_builder.py \
105
+ --usdz-dir /path/to/usdz_dataset \
106
+ --out-dir /path/to/output/usda \
107
+ --template Data/template.usda \
108
+ --usdz-placeholder "@usdz_root[gauss.usda]@" \
109
+ --collision-placeholder "@collision_root@" \
110
+ --usdz-path-template "/path/to/usdz_dataset/{scene_id}.usdz[gauss.usda]" \
111
+ --collision-path-template "/path/to/collision_mesh_dataset/{scene_id}/{scene_id}_collision.usd" \
112
+ --overwrite
113
+ ```
114
+ *Replace `/path/to/usdz_dataset` with the local path to this Hugging Face dataset, and `/path/to/collision_mesh_dataset` with the local path to `spatialverse/SAGE-3D_Collision_Mesh`.*
115
 
116
+ ---
117
 
118
  ## πŸ”— Related Datasets
119
 
120
  This dataset is part of the **SAGE-3D** project:
121
 
122
+ 1. **InteriorGS**: Original 3DGS scenes with semantic annotations
123
+ β†’ [spatialverse/InteriorGS](https://huggingface.co/datasets/spatialverse/InteriorGS)
124
 
125
+ 2. **SAGE-3D InteriorGS USDZ** (This dataset): USDZ format for Isaac Sim
126
+ β†’ [spatialverse/SAGE-3D_InteriorGS_usdz](https://huggingface.co/datasets/spatialverse/SAGE-3D_InteriorGS_usdz)
127
 
128
+ 3. **SAGE-3D Collision Mesh**: Physics-enabled collision bodies
129
+ β†’ [spatialverse/SAGE-3D_Collision_Mesh](https://huggingface.co/datasets/spatialverse/SAGE-3D_Collision_Mesh)
130
 
131
+ 4. **SAGE-3D VLN Data**: Navigation trajectories and instructions
132
+ β†’ [spatialverse/SAGE-3D_VLN_Data](https://huggingface.co/datasets/spatialverse/SAGE-3D_VLN_Data)
133
 
134
  ---
135
 
 
153
  ```bibtex
154
  @misc{miao2025physicallyexecutable3dgaussian,
155
  title={Towards Physically Executable 3D Gaussian for Embodied Navigation},
156
+ author={Bingchen Miao and Rong Wei and Zhiqi Ge and Xiaoquan sun and Shiqi Gao and Jingzhe Zhu and Renhan Wang and Siliang Tang and Jun Xiao and Rui Tang and Juncheng Li},
157
  year={2025},
158
  eprint={2510.21307},
159
  archivePrefix={arXiv},
 
177
 
178
  <div align="center">
179
  <strong>SAGE-3D: Semantically and Physically-Aligned Gaussian Environments for 3D Navigation</strong>
180
+ </div>