https://github.com/autonomousvision/volsurfs
https://github.com/autonomousvision/volsurfs
Last synced: about 1 year ago
JSON representation
- Host: GitHub
- URL: https://github.com/autonomousvision/volsurfs
- Owner: autonomousvision
- License: other
- Created: 2024-08-24T14:21:58.000Z (almost 2 years ago)
- Default Branch: main
- Last Pushed: 2025-05-07T13:45:05.000Z (about 1 year ago)
- Last Synced: 2025-05-07T14:48:06.699Z (about 1 year ago)
- Language: Python
- Size: 120 MB
- Stars: 23
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Volumetric Surfaces (CVPR 2025)
### Representing Fuzzy Geometries with Layered Meshes
### [Project Page](https://autonomousvision.github.io/volsurfs/) | [Paper](https://arxiv.org/pdf/2409.02482) | [Web Demo](https://autonomousvision.github.io/volsurfs/viewer/)
[Stefano Esposito](https://s-esposito.github.io/)1,
[Anpei Chen](https://apchenstu.github.io/)1,
[Christian Reiser](https://creiser.github.io/)1,
[Samuel Rota Bulò](https://scholar.google.com/citations?user=484sccEAAAAJ&hl=it)2,
[Lorenzo Porzi](https://scholar.google.it/citations?user=vW1gaVEAAAAJ&hl=it)2,
[Katja Schwarz](https://katjaschwarz.github.io/)2,
[Christian Richardt](https://richardt.name/)2,
[Michael Zollhöfer](https://zollhoefer.com/)2,
[Peter Kontschieder](https://scholar.google.co.uk/citations?user=CxbDDRMAAAAJ&hl=en)2,
[Andreas Geiger](https://www.cvlibs.net/)1
1[University of Tübingen](https://uni-tuebingen.de/fakultaeten/mathematisch-naturwissenschaftliche-fakultaet/fachbereiche/informatik/lehrstuehle/autonomous-vision/home/), 2Meta Reality Labs
## 🛠️ Installation
```bash
# Clone the repository with submodules
git clone --recursive https://github.com/autonomousvision/volsurfs
cd volsurfs
git submodule update --remote --merge
# Create and activate a conda environment
conda create -n volsurfs python=3.8 cmake=3.31
conda activate volsurfs
# (Optional) Install CUDA toolkit
conda install -c "nvidia/label/cuda-11.8.0" cuda-toolkit
# Install PyTorch with CUDA support
conda install pytorch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 pytorch-cuda=11.8 -c pytorch -c nvidia
# Install Python dependencies
pip install -r requirements.txt
# Install raytracelib
cd submodules/raytracelib
pip install -e .
cd ../..
# Install NVIDIA APEX
cd submodules/apex
pip install . -v --disable-pip-version-check --no-cache-dir --no-build-isolation \
--config-settings "--build-option=--cpp_ext" --config-settings "--build-option=--cuda_ext" ./
cd ../..
# Compile and install VolSurfs
pip install ninja
pip install -e . --no-build-isolation
```
## 📁 Datasets
We use the following datasets for training and evaluation:
- [NeRF-Synthetic](scripts/download/blender.sh)
- [DTU](scripts/download/dtu.sh)
- [Shelly](scripts/download/shelly.sh)
Download scripts are located in `scripts/download/` and will place the datasets in the `data/` directory. You can configure dataset paths in `config/paths_config.cfg`.
Example directory structure:
```
data/
├── shelly/
│ ├── khady/
│ ├── kitten/
│ └── ...
├── dtu/
│ ├── dtu_scan24/
│ ├── dtu_scan37/
│ └── ...
├── blender/
│ ├── lego/
│ └── ...
└── ...
```
## 📈 Reproducing Results
To reproduce the main results (5-Mesh) on the **Shelly** dataset, run:
```bash
bash scripts/train_all_shelly.sh
```
Results will be saved in the `runs/` directory. By default, [Weights & Biases](https://wandb.ai/) logging is enabled; you can disable it in `config/train_config.cfg`.
## 📜 License
This project is licensed under the **Creative Commons Attribution 4.0 International License (CC BY 4.0)**.
See the [LICENSE](LICENSE) file for details.
You are free to use, modify, and distribute this code as long as you provide proper attribution to the original authors.
## 📚 Citation
If you use this work in your research, please consider citing:
```bibtex
@inproceedings{Esposito2025VolSurfs,
author = {Esposito, Stefano and Chen, Anpei and Reiser, Christian and Rota Bulò, Samuel and Porzi, Lorenzo and Schwarz, Katja and Richardt, Christian and Zollhoefer, Michael and Kontschieder, Peter and Geiger, Andreas},
title = {Volumetric Surfaces: Representing Fuzzy Geometries with Layered Meshes},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
year = {2025}
}
```
```bibtex
@misc{Esposito2025MVD,
author = {Esposito, Stefano and Geiger, Andreas},
title = {MVDatasets: Standardized DataLoaders for 3D Computer Vision},
year = {2025},
url = {https://github.com/autonomousvision/mvdatasets},
note = {GitHub repository}
}
```
## 🙏 Acknowledgements
This repository builds upon [Radu Alexandru Rosu](https://radualexandru.github.io/)'s excellent project [permuto_sdf](https://github.com/RaduAlexandru/permuto_sdf).
We thank him for sharing his work and providing a strong foundation.