---
license: apache-2.0
library_name: torch-pointcloud
tags:
- point-cloud
- 3d
- pytorch
- torch-pointcloud
- detr3d
- object-detection
datasets:
- sunrgbd
model-index:
- name: 3detr.sunrgbd.fair
  results:
  - task:
      type: 3d-object-detection
    dataset:
      name: SUN RGB-D
      type: sunrgbd
    metrics:
    - name: mAP@0.25
      type: map
      value: 58.08
    - name: mAP@0.5
      type: map
      value: 29.64
---

# Model card for 3detr.sunrgbd.fair

A 3DETR 3D object detection model (end-to-end set-prediction transformer detector). Trained on SUN RGB-D.

## Model Details

- **Model Type:** 3D object detection
- **Model Stats:**
  - Params (M): 7.3
  - Classes: 10
  - Features: 256
- **Dataset:** SUN RGB-D
- **Metrics:** mAP@0.25 58.08, mAP@0.5 29.64 (reference 58.0)
- **Paper:** [An End-to-End Transformer Model for 3D Object Detection](https://arxiv.org/abs/2109.08141)
- **Converted from:** [facebookresearch/3detr](https://github.com/facebookresearch/3detr) (Apache-2.0)
- **Library:** [torch-pointcloud](https://github.com/arthurdjn/pytorch-pointcloud)

## Install

```bash
pip install torch-pointcloud
```

## Usage

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

model, info = tp.create_model(
    "3detr.sunrgbd.fair",
    task="detection",
    pretrained=True,
    return_info=True,
)
model = model.eval()

# 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,
}
data = info["transform"](sample)
data = collate([data])

with torch.no_grad():
    out = model(data.get("x"), data["pos"], data["batch"])
```

## Feature extraction

```python
with torch.no_grad():
    features = model.forward_features(data.get("x"), data["pos"], data["batch"])  # 256 channels
```

## Citation

```bibtex
@inproceedings{misra2021detr3d,
  title   = {An End-to-End Transformer Model for 3D Object Detection},
  author  = {Ishan Misra and Rohit Girdhar and Armand Joulin},
  booktitle = {ICCV},
  year    = {2021}
}

@inproceedings{song2015sunrgbd,
  title     = {{SUN RGB-D}: A {RGB-D} Scene Understanding Benchmark Suite},
  author    = {Song, Shuran and Lichtenberg, Samuel P. and Xiao, Jianxiong},
  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},
}
```