datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
ModelNet40_Auto_aligned
ModelNet40 Auto Aligned
Auto-aligned version of the ModelNet40 3D CAD dataset. Each sample is an OFF mesh file organized by class and train/test split.
This dataset mirrors the layout of naderalfares/ModelNet40, but uses the auto-aligned meshes from the Princeton ModelNet release.
Dataset structure
modelnet40_auto_aligned/
{class}/
train/{class}_{id}.off
test/{class}_{id}.off
40 classes (airplane, bathtub, bed, …, xbox)
9,843 training meshes
2,468 test… See the full description on the dataset page: https://huggingface.co/datasets/naderalfares/ModelNet40_Auto_aligned.modelnet40modelnet40
Dataset Card for "modelnet40"
More Information needed
modelnet40-2048
Dataset Card for "modelnet40-2048"
More Information needed
ModelNet40modelnet40_normal_resampled-compressedModelNet40
ModelNet40
This is a preprocessed version of the ModelNet40 dataset, created with:
import torch_geometric
from torch_geometric.datasets import ModelNet
from torch_geometric.transforms import Compose, SamplePoints, KNNGraph
torch_geometric.seed.seed_everything(42)
# Standard setup
pre_transform = Compose([
SamplePoints(1024),
KNNGraph(k=16)
])
path = '/mnt/lrmc_pool/ModelNet40'
train_dataset = ModelNet(path, '40', train=True, transform=pre_transform)
test_dataset =… See the full description on the dataset page: https://huggingface.co/datasets/cminst/ModelNet40.modelnet40_multi_viewModelNet40modelnet40_normal_resampled-compressedmodelnet40_normal_resampled-compressedmodelnet40_normal_resampled-compressedModelNet40geotransformer-modelnet40modelnet40