SpectralNeRF: Physically Based Spectral Rendering with Neural Radiance Field

This is the Pytorch implementation of our AAAI 2024 paper SpectralNeRF.

Installation

git clone https://github.com/liru0126/SpectralNeRF.git
cd SpectralNeRF
pip install -r requirements.txt

Data Preparation

The datasets can be downloaded here.

The detailed render configurations for synthetic datasets and the capture instructions for real-world scenes can be found here.

Training

sh train.sh

Testing

sh test.sh

Citation

@inproceedings{li2024spectralnerf,
  title={SpectralNeRF: Physically Based Spectral Rendering with Neural Radiance Field},
  author={Li, Ru and Liu, Jia and Liu, Guanghui and Zhang, Shengping and Zeng, Bing and Liu, Shuaicheng},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={38},
  number={4},
  pages={3154--3162},
  year={2024}
}

Acknowledgments

In this project we use (parts of) the implementations of the following work:

• nerf-pytorch

We thank the respective authors for open sourcing of their implementations.