ViP-NeRF

Official code release accompanying the SIGGRAPH 2023 paper - "ViP-NeRF: Visibility Prior for Sparse Input Neural Radiance Fields". All the published data is available here.

Python Environment

Environment details are available in EnvironmentData/ViP_NeRF_GPU.yml. The environment can be created using conda

conda env create -f ViP_NeRF_GPU.yml

Add the source directory to PYTHONPATH

export PYTHONPATH=<ABSOLUTE_PATH_TO_VIPNERF_DIR>/src:$PYTHONPATH

Set-up Databases

Please follow the instructions in database_utils/README.md file to set up various databases. Instructions for custom databases are also included here.

Generate Priors

Sparse Depth Prior

Please follow the instructions in prior_generators/sparse_depth/README.md file to generate sparse depth prior.

Dense Visibility Prior

Please follow the instructions in prior_generators/visibility/README.md file to generate dense visibility prior.

Training and Inference

The files RealEstateTrainerTester01.py, NerfLlffTrainerTester01.py and DtuTrainerTester01.py contain the code for training, testing and quality assessment along with the configs for the respective databases.

cd src/
python RealEstateTrainerTester01.py
python NerfLlffTrainerTester01.py
python DtuTrainerTester01.py
cd ../

Inference with Pre-trained Models

The train configs are also provided in runs/training/train**** folders for each of the scenes. Please download the trained models from here and place them in the appropriate folders.

Disable the train call in the TrainerTester files and run the respective files. For DTU dataset, we do not provide the pre-trained models for train0044, train0045, train0046. Hence, disable the corresponding inference calls in DtuTrainerTester01.py.

This will run inference using the pre-trained models and also evaluate the synthesized images and reports the performance. To reproduce results from the paper, use the models trained for 50k iterations. For best results, use the models trained for more iterations.

License

MIT License

Copyright (c) 2023 Nagabhushan Somraj, Rajiv Soundararajan

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Citation

If you use this code for your research, please cite our paper

@article{somraj2023VipNeRF,
    title = {{ViP-NeRF}: Visibility Prior for Sparse Input Neural Radiance Fields},
    author = {Somraj, Nagabhushan and Soundararajan, Rajiv},
    booktitle = {ACM Special Interest Group on Computer Graphics and Interactive Techniques (SIGGRAPH)},
    month = {August},
    year = {2023},
    doi = {10.1145/3588432.3591539},
}

If you use outputs/results of ViP-NeRF model in your publication, please specify the version as well. The current version is 1.0.

Acknowledgements

Our code is built on top of NeRFs-Simplified codebase.

For any queries or bugs regarding ViP-NeRF, please raise an issue.