Model card for octformer-base.modelnet40.octree-nn

An OctFormer point cloud classification model (octree-based windowed transformer). Trained on ModelNet40.

Model Details

• Model Type: Point cloud classification
• Model Stats:
  • Params (M): 4.0
  • Input channels: 4
  • Classes: 40
  • Features: 192
• Dataset: ModelNet40
• Metrics: OA 92.02 (reference 92.7)
• Paper: OctFormer: Octree-based Transformers for 3D Point Clouds
• Converted from: octree-nn/octformer (MIT)
• Library: torch-pointcloud

Install

pip install torch-pointcloud

This checkpoint also needs ocnn and dwconv, which need a build matching your torch and CUDA: see the installation guide.

Usage

import torch
import torch_pointcloud as tp
from torch_pointcloud.utils.data import collate

model, info = tp.create_model(
    "octformer-base.modelnet40.octree-nn",
    task="classification",
    pretrained=True,
    return_info=True,
)
model = model.cuda().eval()  # GPU-only kernels

# synthetic sample with the keys a dataset provides
num_points = 8192
sample = {
    "pos": torch.randn(num_points, 3),
    "normal": torch.randn(num_points, 3),
    "face": torch.randint(0, num_points, (2 * num_points, 3)),
}
data = info["transform"](sample)
data = collate([data])
data = {key: value.cuda() for key, value in data.items()}

with torch.no_grad():
    logits = model(data.get("x"), data["octree"], data["octree"].depth)

Feature extraction

with torch.no_grad():
    embeddings = model.forward_features(data.get("x"), data["octree"], data["octree"].depth)

model.reset_classifier(num_classes=0)
with torch.no_grad():
    embeddings = model(data.get("x"), data["octree"], data["octree"].depth)  # (B, 192)

Citation

@article{wang2023octformer,
  title   = {OctFormer: Octree-based Transformers for 3D Point Clouds},
  author  = {Peng-Shuai Wang},
  journal = {ACM Transactions on Graphics},
  volume  = {42},
  number  = {4},
  year    = {2023}
}

@inproceedings{wu2015modelnet,
  title   = {3D ShapeNets: A Deep Representation for Volumetric Shapes},
  author  = {Zhirong Wu and Shuran Song and Aditya Khosla and Fisher Yu and Linguang Zhang and Xiaoou Tang and Jianxiong Xiao},
  booktitle = {CVPR},
  year    = {2015}
}

@software{dujardin2026pytorchpointcloud,
  author  = {Arthur Dujardin},
  title   = {PyTorch PointCloud},
  year    = {2026},
  doi     = {10.5281/zenodo.22159632},
  url     = {https://github.com/arthurdjn/pytorch-pointcloud},
}