Datasets:
Commit
·
0485ae5
1
Parent(s):
44e46d6
Update readme
Browse files- README.md +19 -12
- assets/grippers.png +3 -0
README.md
CHANGED
|
@@ -3,16 +3,17 @@ tags:
|
|
| 3 |
- robotics
|
| 4 |
- grasping
|
| 5 |
- simulation
|
| 6 |
-
|
|
|
|
| 7 |
---
|
| 8 |
|
| 9 |
-
# GraspGen: Scaling
|
| 10 |
-
GraspGen is a large-scale simulated grasp dataset for multiple robot embodiments and grippers
|
| 11 |
|
| 12 |
<img src="assets/cover.png" width="1000" height="250" title="readme1">
|
| 13 |
|
| 14 |
|
| 15 |
-
We release over 57 million grasps, computed for a subset of 8515 objects from the [Objaverse XL](https://objaverse.allenai.org/) (LVIS) dataset.
|
| 16 |
|
| 17 |
<img src="assets/montage2.png" width="1000" height="500" title="readme2">
|
| 18 |
|
|
@@ -28,20 +29,23 @@ splits/
|
|
| 28 |
robotiq2f140/{train/valid}_scenes.json
|
| 29 |
suction/{train/valid}_scenes.json
|
| 30 |
```
|
| 31 |
-
We release test-train splits along with the grasp dataset.
|
| 32 |
|
| 33 |
Each json file in the shard has the following data in a python dictionary. Note that `num_grasps=2000` per object.
|
| 34 |
```
|
| 35 |
‘object’/
|
| 36 |
-
‘scale’ # This is the scale of the asset
|
| 37 |
‘grasps’/
|
| 38 |
‘object_in_gripper’ # boolean mask indicating grasp success, [num_grasps X 1]
|
| 39 |
‘transforms’ # Pose of the gripper in homogenous matrices, [num_grasps X 4 X 4]
|
| 40 |
```
|
| 41 |
|
|
|
|
|
|
|
|
|
|
| 42 |
## Visualizing the dataset
|
| 43 |
|
| 44 |
-
We have provided some standalone scripts for visualizing this dataset. See the header of the [visualize_dataset.py](scripts/visualize_dataset.py) for installation instructions
|
| 45 |
|
| 46 |
Before running any of the visualization scripts, remember to start meshcat-server in a separate terminal:
|
| 47 |
``` shell
|
|
@@ -53,18 +57,21 @@ To visualize a single object from the dataset, alongside its grasps:
|
|
| 53 |
cd scripts/ && python visualize_dataset.py --dataset_path /path/to/dataset --object_uuid {object_uuid} --object_file /path/to/mesh --gripper_name {choose from: franka, suction, robotiq2f140}
|
| 54 |
```
|
| 55 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 56 |
## Objaverse dataset
|
| 57 |
Please download the Objaverse XL (LVIS) objects separately. See the helper script [download_objaverse.py](scripts/download_objaverse.py) for instructions and usage.
|
|
|
|
| 58 |
|
| 59 |
## License
|
| 60 |
License Copyright © 2025, NVIDIA Corporation & affiliates. All rights reserved.
|
| 61 |
|
| 62 |
-
|
| 63 |
-
The dataset is released under a CC-BY 4.0 License.
|
| 64 |
-
|
| 65 |
-
The visualization code is released under the [NVIDIA source code license](LICENSE).
|
| 66 |
|
| 67 |
## Contact
|
| 68 |
|
| 69 |
-
Please reach out to [Adithya Murali](adithyamurali.com) ([email protected]) and [Clemens Eppner](https://clemense.github.io/) ([email protected]) for further enquiries
|
| 70 |
|
|
|
|
| 3 |
- robotics
|
| 4 |
- grasping
|
| 5 |
- simulation
|
| 6 |
+
- nvidia
|
| 7 |
+
license: "cc-by-nc-4.0"
|
| 8 |
---
|
| 9 |
|
| 10 |
+
# GraspGen: Scaling Sim2Real Grasping
|
| 11 |
+
GraspGen is a large-scale simulated grasp dataset for multiple robot embodiments and grippers.
|
| 12 |
|
| 13 |
<img src="assets/cover.png" width="1000" height="250" title="readme1">
|
| 14 |
|
| 15 |
|
| 16 |
+
We release over 57 million grasps, computed for a subset of 8515 objects from the [Objaverse XL](https://objaverse.allenai.org/) (LVIS) dataset. These grasps are specific to three grippers: Franka Panda, the Robotiq-2f-140 industrial gripper, and a single-contact suction gripper (30mm radius).
|
| 17 |
|
| 18 |
<img src="assets/montage2.png" width="1000" height="500" title="readme2">
|
| 19 |
|
|
|
|
| 29 |
robotiq2f140/{train/valid}_scenes.json
|
| 30 |
suction/{train/valid}_scenes.json
|
| 31 |
```
|
| 32 |
+
We release test-train splits along with the grasp dataset. The splits are made randomly based on object instances.
|
| 33 |
|
| 34 |
Each json file in the shard has the following data in a python dictionary. Note that `num_grasps=2000` per object.
|
| 35 |
```
|
| 36 |
‘object’/
|
| 37 |
+
‘scale’ # This is the scale of the asset, float
|
| 38 |
‘grasps’/
|
| 39 |
‘object_in_gripper’ # boolean mask indicating grasp success, [num_grasps X 1]
|
| 40 |
‘transforms’ # Pose of the gripper in homogenous matrices, [num_grasps X 4 X 4]
|
| 41 |
```
|
| 42 |
|
| 43 |
+
The coordinate frame convention for the three grippers are provided below:
|
| 44 |
+
<img src="assets/grippers.png" width="450" height="220" title="readme3">
|
| 45 |
+
|
| 46 |
## Visualizing the dataset
|
| 47 |
|
| 48 |
+
We have provided some minimal, standalone scripts for visualizing this dataset. See the header of the [visualize_dataset.py](scripts/visualize_dataset.py) for installation instructions.
|
| 49 |
|
| 50 |
Before running any of the visualization scripts, remember to start meshcat-server in a separate terminal:
|
| 51 |
``` shell
|
|
|
|
| 57 |
cd scripts/ && python visualize_dataset.py --dataset_path /path/to/dataset --object_uuid {object_uuid} --object_file /path/to/mesh --gripper_name {choose from: franka, suction, robotiq2f140}
|
| 58 |
```
|
| 59 |
|
| 60 |
+
To sequentially visualize a list of objects with its grasps:
|
| 61 |
+
```shell
|
| 62 |
+
cd scripts/ && python visualize_dataset.py --dataset_path /path/to/dataset --uuid_list {path to a splits.json file} --uuid_object_paths_file {path to a json file mapping uuid to absolute path of meshes} --gripper_name {choose from: franka, suction, robotiq2f140}
|
| 63 |
+
```
|
| 64 |
+
|
| 65 |
## Objaverse dataset
|
| 66 |
Please download the Objaverse XL (LVIS) objects separately. See the helper script [download_objaverse.py](scripts/download_objaverse.py) for instructions and usage.
|
| 67 |
+
Note that running this script autogenerates a file that maps from `UUID` to the asset mesh path, which you can pass in as input `uuid_object_paths_file` to the `visualize_dataset.py` script.
|
| 68 |
|
| 69 |
## License
|
| 70 |
License Copyright © 2025, NVIDIA Corporation & affiliates. All rights reserved.
|
| 71 |
|
| 72 |
+
Both the dataset and visualization code is released under a CC-BY-NC 4.0 [License](LICENSE_DATASET).
|
|
|
|
|
|
|
|
|
|
| 73 |
|
| 74 |
## Contact
|
| 75 |
|
| 76 |
+
Please reach out to [Adithya Murali](http://adithyamurali.com) ([email protected]) and [Clemens Eppner](https://clemense.github.io/) ([email protected]) for further enquiries.
|
| 77 |
|
assets/grippers.png
ADDED
|
Git LFS Details
|