donglaix commited on
Commit
abab30f
·
verified ·
1 Parent(s): ccf24c1

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -175,7 +175,8 @@ v1.0 - Initial research release with Geometry Transformer (85.81M parameters + 3
175
  Training set contains 1,333 objects with 6,477 segments and 28.7M voxels. Data modalities include (i) 3D meshes (triangle meshes in USD format with part-level segmentation), multi-view RGB images (512×512 rendered views), text (English material names), and tabular material properties (E, ν, ρ triplets); (ii) Nature of content is non-personal, proprietary 3D assets (NVIDIA asset libraries) combined with public domain material science data, no copyright-protected creative content, machine-generated annotations (VLM) constrained by human-measured physical properties; (iii) Linguistic characteristics include English material names and semantic object labels. No sensors were used for data collection; 3D assets are human-modeled, material properties are from laboratory measurements (ASTM standard testing), and images are path-traced renders. Average 4.86 segments per object (±11.97 std dev), 21,537 voxels per object (±23,431 std dev). Material property ranges: E [1.0×10^5, 2.8×10^11 Pa], ν [0.16, 0.49], ρ [50, 19,300 kg/m³].
176
 
177
  ## Testing Dataset:
178
- **Link:** GVM test split - Internal NVIDIA dataset, ABO-500 - [Amazon Berkeley Objects](https://amazon-berkeley-objects.s3.amazonaws.com/index.html)
 
179
 
180
  **Data Collection Method by dataset:**
181
  * Hybrid: Human, Automated (VLM-assisted)
@@ -187,7 +188,6 @@ Training set contains 1,333 objects with 6,477 segments and 28.7M voxels. Data m
187
  Test set contains 166 objects with 1,060 segments and 4.9M voxels (13.1% of total dataset). Same modalities, content nature, and linguistic characteristics as training data. Average 6.39 segments per object (±11.33 std dev), 29,571 voxels per object (±25,987 std dev). Held-out objects ensure no overlap with training data.
188
 
189
  ## Evaluation Dataset:
190
- **Link:**
191
  - Primary: GVM test split (166 objects, 4.9M voxel annotations)
192
  - Secondary: [ABO-500](https://amazon-berkeley-objects.s3.amazonaws.com/index.html) mass estimation benchmark - 500 objects with ground truth mass labels
193
 
 
175
  Training set contains 1,333 objects with 6,477 segments and 28.7M voxels. Data modalities include (i) 3D meshes (triangle meshes in USD format with part-level segmentation), multi-view RGB images (512×512 rendered views), text (English material names), and tabular material properties (E, ν, ρ triplets); (ii) Nature of content is non-personal, proprietary 3D assets (NVIDIA asset libraries) combined with public domain material science data, no copyright-protected creative content, machine-generated annotations (VLM) constrained by human-measured physical properties; (iii) Linguistic characteristics include English material names and semantic object labels. No sensors were used for data collection; 3D assets are human-modeled, material properties are from laboratory measurements (ASTM standard testing), and images are path-traced renders. Average 4.86 segments per object (±11.97 std dev), 21,537 voxels per object (±23,431 std dev). Material property ranges: E [1.0×10^5, 2.8×10^11 Pa], ν [0.16, 0.49], ρ [50, 19,300 kg/m³].
176
 
177
  ## Testing Dataset:
178
+ - GVM test split
179
+ - ABO-500 - [Amazon Berkeley Objects](https://amazon-berkeley-objects.s3.amazonaws.com/index.html)
180
 
181
  **Data Collection Method by dataset:**
182
  * Hybrid: Human, Automated (VLM-assisted)
 
188
  Test set contains 166 objects with 1,060 segments and 4.9M voxels (13.1% of total dataset). Same modalities, content nature, and linguistic characteristics as training data. Average 6.39 segments per object (±11.33 std dev), 29,571 voxels per object (±25,987 std dev). Held-out objects ensure no overlap with training data.
189
 
190
  ## Evaluation Dataset:
 
191
  - Primary: GVM test split (166 objects, 4.9M voxel annotations)
192
  - Secondary: [ABO-500](https://amazon-berkeley-objects.s3.amazonaws.com/index.html) mass estimation benchmark - 500 objects with ground truth mass labels
193