Model card for fcaf3d.s3dis-area5.mmdet3d

A FCAF3D 3D object detection model (fully convolutional anchor-free sparse voxel detector). Trained on S3DIS (Area 5).

Model Details

• Model Type: 3D object detection
• Model Stats:
  • Params (M): 70.5
  • Input channels: 3
  • Classes: 5
  • Features: 512
• Dataset: S3DIS (Area 5)
• Metrics: mAP@0.25 66.31, mAP@0.5 46.71
• Paper: FCAF3D: Fully Convolutional Anchor-Free 3D Object Detection
• Converted from: open-mmlab/mmdetection3d (Apache-2.0)
• 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(
    "fcaf3d.s3dis-area5.mmdet3d",
    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():
    out = model(data.get("x"), data["pos"], data["batch"])

Feature extraction

with torch.no_grad():
    features = model.forward_features(data.get("x"), data["pos"], data["batch"])  # 512 channels

Citation

@inproceedings{rukhovich2022fcaf3d,
  title   = {{FCAF3D}: Fully Convolutional Anchor-Free 3D Object Detection},
  author  = {Danila Rukhovich and Anna Vorontsova and Anton Konushin},
  booktitle = {ECCV},
  year    = {2022}
}

@inproceedings{armeni2016s3dis,
  title     = {{3D} Semantic Parsing of Large-Scale Indoor Spaces},
  author    = {Armeni, Iro and Sener, Ozan and Zamir, Amir R. and Jiang, Helen and Brilakis, Ioannis and Fischer, Martin and Savarese, Silvio},
  booktitle = {CVPR},
  year      = {2016}
}

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