This repository is built for the paper:

Learning Geometry-Disentangled Representation for Complementary Understanding of 3D Object Point Cloud (AAAI2021) [arXiv]
by Mutian Xu*, Junhao Zhang*, Zhipeng Zhou, Mingye Xu, Xiaojuan Qi and Yu Qiao.

If you find the code or trained models useful, please consider citing:

@inproceedings{xu2021gdanet,
  title={Learning Geometry-Disentangled Representation for Complementary Understanding of 3D Object Point Cloud}, 
  author={Mutian Xu and Junhao Zhang and Zhipeng Zhou and Mingye Xu and Xiaojuan Qi and Yu Qiao},
  booktitle={AAAI},
  year={2021}
}

20/05/2022:

GDANet becomes a standard and strong baseline model for robust 3D recognition, it gains competitive performance on OmniObject3D, ModelNet-C and ModelNet40-C datasets for object classification under corruptions.

• Linux (tested on Ubuntu 14.04/16.04)
• Python 3.5+
• PyTorch 1.0+

• Create the folder to symlink the data later:

  mkdir -p data

• Object Classification:

  Download and unzip ModelNet40 (415M), then symlink the path to it as follows (you can alternatively modify the path here) :

  ln -s /path to modelnet40/modelnet40_ply_hdf5_2048 data

• Shape Part Segmentation:

  Download and unzip ShapeNet Part (674M), then symlink the path to it as follows (you can alternatively modify the path here) :

  ln -s /path to shapenet part/shapenetcore_partanno_segmentation_benchmark_v0_normal data

• Train:

  python main_cls.py

• Test:

  • Run the voting evaluation script, after this voting you will get an accuracy of 93.8% if all things go right:

    python voting_eval_modelnet.py --model_path 'pretrained/GDANet_ModelNet40_93.4.t7'

  • You can also directly evaluate our pretrained model without voting to get an accuracy of 93.4%:

    python main.py --eval True --model_path 'pretrained/GDANet_ModelNet40_93.4.t7'

• Train:

  • Training from scratch:

    python main_ptseg.py

  • If you want resume training from checkpoints, specify resume in the args:

    python main_ptseg.py --resume True

• Test:

  You can choose to test the model with the best instance mIoU, class mIoU or accuracy, by specifying eval and model_type in the args:

  • python main_ptseg.py --eval True --model_type 'insiou' (best instance mIoU, default)

  • python main_ptseg.py --eval True --model_type 'clsiou' (best class mIoU)

  • python main_ptseg.py --eval True --model_type 'acc' (best accuracy)

  Note: This works only after you trained the model or if you have the checkpoint in checkpoints/GDANet. If you run the training from scratch the checkpoints will automatically be generated there.

The following tables report the current performances on different tasks and datasets.

Please contact Mutian Xu ([email protected]) or Junhao Zhang ([email protected]) for further discussion.

This code is is partially borrowed from DGCNN and PointNet++.