README.md

MuRF: Multi-Baseline Radiance Fields

Haofei Xu · Anpei Chen · Yuedong Chen · Christos Sakaridis · Yulun Zhang
Marc Pollefeys · Andreas Geiger · Fisher Yu

CVPR 2024

Paper | Project Page

murf_teaser_video.mp4

MuRF supports multiple different baseline settings.

MuRF achieves state-of-the-art performance under various evaluation settings.

Installation

Our code is developed based on pytorch 1.10.1, CUDA 11.3 and python 3.8.

We recommend using conda for installation:

conda create -n murf python=3.8
conda activate murf
pip install -r requirements.txt

Model Zoo

The models are hosted on Hugging Face 🤗 : https://huggingface.co/haofeixu/murf

Model details can be found at MODEL_ZOO.md.

Datasets

The datasets used to train and evaluate our models are detailed in DATASETS.md

Evaluation

The evaluation scripts used to reproduce the numbers in our paper are detailed in scripts/*_evaluate.sh.

Rendering

The rendering scripts are detailed in scripts/*_render.sh.

Training

The training scripts are detailed in scripts/*_train.sh.

Citation

@inproceedings{xu2024murf,
      title={MuRF: Multi-Baseline Radiance Fields},
      author={Xu, Haofei and Chen, Anpei and Chen, Yuedong and Sakaridis, Christos and Zhang, Yulun and Pollefeys, Marc and Geiger, Andreas and Yu, Fisher},
      booktitle={CVPR},
      year={2024}
    }

Acknowledgements

This repo is heavily based on MatchNeRF, thanks Yuedong Chen for this fantastic work. This project also borrows code from several other repos: GMFlow, UniMatch, latent-diffusion, MVSNeRF, IBRNet, ENeRF and cross_attention_renderer. We thank the original authors for their excellent work.