Datasets:

Modalities:
3D
Languages:
English
Tags:
3d
DOI:
nbanerje commited on
Commit
3a715c4
·
1 Parent(s): 23012d7

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +22 -2
README.md CHANGED
@@ -9,7 +9,19 @@ tags:
9
  ---
10
  DATASET DESCRIPTION:
11
 
12
- The Geometric Breaks dataset consists of 3D models of 25,249 broken object models and their repair parts. The broken models are generated by subtracting fracturing shapes from a total of 22,165 3D complete object models. Complete object models have been acquired from the Google Scanned Object (GSO) dataset (https://app.gazebosim.org/GoogleResearch/fuel/collections/Scanned%20Objects%20by%20Google%20Research), and 8 classes in the ShapeNetCore.v2 dataset (https://huggingface.co/datasets/ShapeNet/ShapeNetCore). To generate the dataset, we have waterproofed complete meshes, normalized them to a [-0.5,0.5]^3 cube, and generated broken shapes and repair counteparts by conducting Boolean subtraction with a fracturing shape from the waterproofed complete mesh. Fracturing shapes come from one of three geometric primitives: cube, icosphere, and subivided icosphere, randomly translated and rotated to generate breaks over various parts of the original model. Vertices on the primitive surface are randomly perturbed to simulate roughness. Since the subtraction process generates multiple fragments, the dataset also contains individual 3D models of the fragmented parts. Models from following ShapeNetCore.v2 classes are used: Jar (03593526), Bottle (02876657), Mug (03797390), Airplane (02691156), Chair (03001627), Car (02958343), Sofa (04256520), Table (04379243).
 
 
 
 
 
 
 
 
 
 
 
 
13
 
14
  DATASET STRUCTURE:
15
  The dataset is shared as the following ZIP files:
@@ -19,7 +31,15 @@ The dataset is shared as the following ZIP files:
19
  - vehicles.zip: airplane and car objects
20
  - GSO.zip: GSO objects
21
 
22
- Each zip file when unzipped generates the folders corresponding to its object classes. For GSO.zip, a folder called GSO is created. For the remaining zip files, folders are created labeled by the the ShapeNetCore IDs of the 8 classes used in the dataset. Within each class folder, object models are stored in the subfolder hierarchy '[classID]/[objectID]/models/'. The value of '[classID]' corresponds to the ShapeNetCore ID or 'GSO'. The value of '[objectID]' corresponds to the ShapeNetCore object model ID for ShapeNetCore models and to the object model descriptor name with underscores as used to name object models in GSO. The use of the same ShapeNetCore class IDs, and ShapeNetCore or GSO object IDs enables one-to-one cross-referencing with the original ShapeNetCore and GSO datasets.
 
 
 
 
 
 
 
 
23
 
24
  Within each '[classID]/[objectID]/models' subfolder, models are stored as triangular meshes in the PLY file format (https://gamma.cs.unc.edu/POWERPLANT/papers/ply.pdf) as follows:
25
  - The waterproofed version of the original model is provided in the file 'model_c.ply', where 'c' stands for complete.
 
9
  ---
10
  DATASET DESCRIPTION:
11
 
12
+ The Geometric Breaks dataset consists of 3D models of 25,249 broken object models and their repair parts.
13
+ The broken models are generated by subtracting fracturing shapes from a total of 22,165 3D complete object models.
14
+ Complete object models have been acquired from the Google Scanned Object (GSO) dataset
15
+ (https://app.gazebosim.org/GoogleResearch/fuel/collections/Scanned%20Objects%20by%20Google%20Research), and 8 classes in
16
+ the ShapeNetCore.v2 dataset (https://huggingface.co/datasets/ShapeNet/ShapeNetCore).
17
+ To generate the dataset, we have waterproofed complete meshes, normalized them to a [-0.5,0.5]^3 cube, and
18
+ generated broken shapes and repair counteparts by conducting Boolean subtraction with a fracturing shape from the
19
+ waterproofed complete mesh. Fracturing shapes come from one of three geometric primitives: cube, icosphere, and subivided icosphere,
20
+ randomly translated and rotated to generate breaks over various parts of the original model.
21
+ Vertices on the primitive surface are randomly perturbed to simulate roughness.
22
+ Since the subtraction process generates multiple fragments, the dataset also contains individual 3D models of the fragmented parts.
23
+ Models from following ShapeNetCore.v2 classes are used:
24
+ Jar (03593526), Bottle (02876657), Mug (03797390), Airplane (02691156), Chair (03001627), Car (02958343), Sofa (04256520), Table (04379243).
25
 
26
  DATASET STRUCTURE:
27
  The dataset is shared as the following ZIP files:
 
31
  - vehicles.zip: airplane and car objects
32
  - GSO.zip: GSO objects
33
 
34
+ Each zip file when unzipped generates the folders corresponding to its object classes.
35
+ For GSO.zip, a folder called GSO is created.
36
+ For the remaining zip files, folders are created labeled by the the ShapeNetCore IDs of the 8 classes used in the dataset.
37
+ Within each class folder, object models are stored in the subfolder hierarchy '[classID]/[objectID]/models/'.
38
+ The value of '[classID]' corresponds to the ShapeNetCore ID or 'GSO'.
39
+ The value of '[objectID]' corresponds to the ShapeNetCore object model ID for ShapeNetCore models and
40
+ to the object model descriptor name with underscores as used to name object models in GSO.
41
+ The use of the same ShapeNetCore class IDs, and ShapeNetCore or GSO object IDs enables
42
+ one-to-one cross-referencing with the original ShapeNetCore and GSO datasets.
43
 
44
  Within each '[classID]/[objectID]/models' subfolder, models are stored as triangular meshes in the PLY file format (https://gamma.cs.unc.edu/POWERPLANT/papers/ply.pdf) as follows:
45
  - The waterproofed version of the original model is provided in the file 'model_c.ply', where 'c' stands for complete.