αSurf: Implicit Surface Reconstruction for Semi-Transparent and Thin Objects with Decoupled Geometry and Opacity

Project Page, Data

Setup

First create the virtualenv; we recommend using conda:

conda env create -f environment.yml
conda activate alphasurf

Then install pytorch with CUDA support via:

conda install pytorch==1.11.0 torchvision==0.12.0 torchaudio==0.11.0 cudatoolkit=11.3 -c pytorch

Then install the C++/CUDA extension at root by simply running

pip install .

In the repo root directory.

If your CUDA toolkit is older than 11, then you will need to install CUB as follows: conda install -c bottler nvidiacub. Since CUDA 11, CUB is shipped with the toolkit.

Train and eval

An example script for training Plenoxels + our method is avaliable at train_eval.sh

bash train_eval.sh

Citation:

@inproceedings{alphasurf,
    title={AlphaSurf: Implicit Surface Reconstruction for Semi-Transparent and Thin Objects with Decoupled Geometry and Opacity},
    author={Tianhao Walter Wu and Hanxue Liang and Fangcheng Zhong and Gernot Riegler and Shimon Vainer and Jiankang Deng and Cengiz Oztireli}
    journal={3DV},
    year={2025}
}