https://github.com/somtambe/modelnet10-dataset
Compressed really aesthetic handling of the humongous ModelNet10 3D Vision dataset
https://github.com/somtambe/modelnet10-dataset
compressed datasets modelnet-dataset
Last synced: about 1 year ago
JSON representation
Compressed really aesthetic handling of the humongous ModelNet10 3D Vision dataset
- Host: GitHub
- URL: https://github.com/somtambe/modelnet10-dataset
- Owner: SomTambe
- Created: 2020-05-16T04:53:46.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2020-05-21T10:06:13.000Z (about 6 years ago)
- Last Synced: 2025-04-05T11:41:27.068Z (about 1 year ago)
- Topics: compressed, datasets, modelnet-dataset
- Language: Python
- Homepage:
- Size: 10.8 MB
- Stars: 5
- Watchers: 1
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# ModelNet10-dataset
Compressed really aesthetic handling of the humongous ModelNet10 3D Vision dataset
This contains the numpy array file of the huge ModelNet10 dataset.
Original dataset size ~ 2.2 Gigabytes
Numpy file size ~ 8 Gigabytes
Compressed file size ~ 32 Megabytes
This can all be seen due to the very bad handling of binvox files, and on the top of that numpy arrays.
Please star this repo if you find this useful :grin:
## Usage example -
```python
import gzip
import numpy as np
with gzip.open('modelnet10.npy.gz','rb') as f:
arr=np.load(f)
print(type(arr),arr.shape)
```
Output -
```python
(3992, 1, 64, 64, 64)
```
Let's check if it works -
```python
import matplotlib.pyplot as plt
fig = plt.figure()
ax = fig.gca(projection='3d')
ax.voxels(arr[78][0],facecolors='red')
plt.show()
```
Output -
## Scripts to generate from source
If you have raw OFF files, what you can do is go to the scripts folder, and look at the scripts I used to get to the final file format.
I know they seem a joke, but I did not have the time to revitalize them, so make the changes accordingly :grin: .