This repository contains an implementation to the ICML 2020 paper: "Implicit Geometric Regualrization for Learning Shapes".

IGR is a deep learning approach for learning implicit signed distance representations directly from raw point clouds with or without normal data. Our method aims to find an SDF by optimizing the network to solve the eikonal equation with the input point cloud as boundary condition. Although this is an ill posed condition we enjoy an implicit regualrization coming from the optimization procedure itself which aims our method to simple natrual solutions as can be seen on the figure above.

For more details:

paper: https://arxiv.org/abs/2002.10099.

video: https://youtu.be/6cOvBGBQF9g.

The code is compatible with python 3.7 and pytorch 1.2. In addition, the following packages are required:
numpy, pyhocon, plotly, scikit-image, trimesh.

IGR can be used to reconstruct a single surface given a point cloud with or without normal data. Adjust reconstruction/setup.json to the path of the input 2D/3D point cloud:

train
{
  ...
  d_in=D
  ...
  input_path = your_path
  ...
}

Where D=3 in case we use 3D data or 2 if we use 2D. We support xyz,npy,npz,ply files.

Then, run training:

cd ./code
python reconstruction/run.py 

Finally, to produce the meshed surface, run:

cd ./code
python reconstruction/run.py --eval --checkpoint CHECKPOINT

where CHECKPOINT is the epoch you wish to evaluate of 'latest' if you wish to take the most recent epoch.

The raw scans can be downloaded from http://dfaust.is.tue.mpg.de/downloads. In order to sample point clouds with normals use:

cd ./code
python preprocess/dfaust.py --src-path SRC_PATH --out-path OUT_PATH

where SRC_PATH is the absoule path of the directory with the original D-Faust scans, and OUT_PATH is the absolute path of the directory on which you wish to output the processed point clouds.

It is also possible to process only train data with input --mode 0 our only test with --mode 1.

In case you wish to only process part of the data (e.g. for parallel processing) it is possible by adding --names NAME_1,NAME_2,...,NAME_k where NAME_i is one of D-Faust shapes e.g. 50002, 50020.

After preprocessing ended adjust the file ./shapespace/dfaust_setup.conf to the cur path of the data:

train
{
  ...
  dataset_path = OUT_PATH/dfaust_processed
  ...
}

We have uploaded IGR trained network. To produce predictions on unseen test scans, run:

cd ./code
python shapespace/eval.py --checkpoint 1200 --exp-name dfaust_pretrained --split dfaust/test_all.json --exps-dir trained_models

In case you wish to generate less models you can use --split dfaust/test_models.json

To meshs of latent interpolation between two shapes use:

cd ./code
python shapespace/interpolate.py --interval INTERVAL --checkpoint 1200 --exp-name dfaust_pretrained --exps-dir trained_models

Where INTERVAL is the number (int) linspace of latent interpolations.

In case you wish to interpolate different shapes adjust the file dfaust/interpolation.json

If you want to train IGR yourself, run:

cd ./code
python shapespace/train.py

If you find our work useful in your research, please consider citing:

@incollection{icml2020_2086,
 author = {Gropp, Amos and Yariv, Lior and Haim, Niv and Atzmon, Matan and Lipman, Yaron},
 booktitle = {Proceedings of Machine Learning and Systems 2020},
 pages = {3569--3579},
 title = {Implicit Geometric Regularization for Learning Shapes},
 year = {2020}
}

• Yariv et al. - Multiview Neural Surface Reconstruction with Implicit Lighting and Material
• Atzmon & Lipman. - SAL++: Sign Agnostic Learning with Derivatives (2020)
• Atzmon & Lipman. - SAL: Sign Agnostic Learning of Shapes From Raw Data (2020)
• Atzmon et al. - Controlling Neural Level Sets (2019)