Install

git clone https://github.com/vc-bonn/neu-pig.git neu-pig
cd neu-pig

git submodule update --init

conda create -n neupig python=3.12
conda activate neupig


conda install -y -c conda-forge openblas cmake ninja
pip install torch==2.6.0 torchvision==0.21.0 torchaudio==2.6.0 --index-url https://download.pytorch.org/whl/cu126
pip install ext/pcgrid
pip install tqdm tensorboard scikit-learn charonload cmake gpytoolbox imageio matplotlib ninja open3d opencv-python pykdtree trimesh pymeshlab openmesh tensorboardx gitpython rich
pip install --no-build-isolation "git+https://github.com/facebookresearch/pytorch3d.git"

Main.py Arguments

python Main.py --...
[-m / --methodConfig] - Path to the method config file (str)
[-se / --seed] - random seed (int)
[-d / --device] - cuda device number (int)
[-t / --target] - which target type to use. If pcl is chosen, make sure that the directory path is as configured in the AMA dataset. ("pcl"/"obj")
[-np / --number_points] - if [-t == "obj"] set the number of target points sampled per obj file (int)
[-o / --out_path] - output directory (str)
[-dp / --directory_path] - path to the input objects, see PATH-STRUCTURE section (str)
[-s / --skip] - how many objects to skip (int)
[-ns / --noise] - standard deviation of gaussian noise added to target points (float)

PATH-STRUCTURE

To download the preprocessed data, see the data section of Dynosurf.

--AMA
|
|--crane_0010
|-|
|-|--gt
|-|--pcl_seqs
|-|--points_clouds
|
|--crane_0027
|-|
|-|--gt
|-|--pcl_seqs
|-|--points_clouds

RUNS

All run configs are predefined in configs/method/runs. For the ablation study see the configs/method/ablations path. Alter the path arguments accordingly [-o / -dp]. To create the extended AMA dataset, with sequences in the range of 40-120 frames, see src/io/dataset/process_ama_dataset.py.

For the chained execution we supply a run.py script, see src/run.py. See the --help arguments for all the run/supplemental run options. We support multi-gpu runs, setting multiple gpu ids via --devices.

To run the initialization methods from DynoSurf or Motion2VecSets, make sure to install their required packages. For Motion2VecSets, we recommend creating a separate conda environment with the appropriate Python version to avoid version conflicts. For the execution of DynoSurf, please refer to their repository, or execute src/related_work/run_dynosurf.py. Paths need to be altered according to your setup.