Model card for spformer-unet.scannet20.open3d

A SPFormer-UNet point cloud semantic segmentation model (sparse convolutional U-Net from SPFormer). Trained on ScanNet (20 classes).

Model Details

• Model Type: Point cloud semantic segmentation
• Model Stats:
  • Params (M): 17.3
  • Input channels: 3
  • Classes: 20
  • Features: 32
• Dataset: ScanNet (20 classes)
• Metrics: mIoU 67.71, OA 87.99
• Paper: Superpoint Transformer for 3D Scene Instance Segmentation
• Converted from: isl-org/Open3D-ML (MIT)
• Library: torch-pointcloud

Install

pip install torch-pointcloud

This checkpoint also needs spconv, which needs 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(
    "spformer-unet.scannet20.open3d",
    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),
    "color": torch.rand(num_points, 3) * 255,
    "normal": torch.randn(num_points, 3),
    "segment": torch.zeros(num_points, dtype=torch.long),
    "instance": torch.zeros(num_points, dtype=torch.long),
}
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["pos_grid"], data["batch"])

Feature extraction

with torch.no_grad():
    features = model.forward_features(data.get("x"), data["pos_grid"], data["batch"])

model.reset_classifier(num_classes=0)
with torch.no_grad():
    features = model(data.get("x"), data["pos_grid"], data["batch"])  # (N, 32)

Citation

@inproceedings{sun2023spformer,
  title   = {Superpoint Transformer for 3D Scene Instance Segmentation},
  author  = {Jiahao Sun and Chunmei Qing and Junpeng Tan and Xiangmin Xu},
  booktitle = {AAAI},
  year    = {2023}
}

@inproceedings{dai2017scannet,
  title   = {ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes},
  author  = {Angela Dai and Angel X. Chang and Manolis Savva and Maciej Halber and Thomas Funkhouser and Matthias Nießner},
  booktitle = {CVPR},
  year    = {2017}
}

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