D-NeRF: Neural Radiance Fields for Dynamic Scenes

[Project] [Paper]

D-NeRF is a method for synthesizing novel views, at an arbitrary point in time, of dynamic scenes with complex non-rigid geometries. We optimize an underlying deformable volumetric function from a sparse set of input monocular views without the need of ground-truth geometry nor multi-view images.

This project is an extension of NeRF enabling it to model dynmaic scenes. The code heavily relays on NeRF-pytorch.

Installation

git clone https://github.com/albertpumarola/D-NeRF.git
cd D-NeRF
conda create -n dnerf python=3.6
conda activate dnerf
pip install -r requirements.txt
cd torchsearchsorted
pip install .
cd ..

Download Pre-trained Weights

You can download the pre-trained models from drive or dropbox. Unzip the downloaded data to the project root dir in order to test it later. See the following directory structure for an example:

├── logs 
│   ├── mutant
│   ├── standup 
│   ├── ...

Download Dataset

You can download the datasets from drive or dropbox. Unzip the downloaded data to the project root dir in order to train. See the following directory structure for an example:

├── data 
│   ├── mutant
│   ├── standup 
│   ├── ...

Demo

We provide simple jupyter notebooks to explore the model. To use them first download the pre-trained weights and dataset.

Description	Jupyter Notebook
Synthesize novel views at an arbitrary point in time.	render.ipynb
Reconstruct mesh at an arbitrary point in time.	reconstruct.ipynb
Quantitatively evaluate trained model.	metrics.ipynb

Test

First download pre-trained weights and dataset. Then,

python run_dnerf.py --config configs/mutant.txt --render_only --render_test

This command will run the mutant experiment. When finished, results are saved to ./logs/mutant/renderonly_test_799999 To quantitatively evaluate model run metrics.ipynb notebook

Train

First download the dataset. Then,

conda activate dnerf
export PYTHONPATH='path/to/D-NeRF'
export CUDA_VISIBLE_DEVICES=0
python run_dnerf.py --config configs/mutant.txt

Citation

If you use this code or ideas from the paper for your research, please cite our paper:

@article{pumarola2020d,
  title={D-NeRF: Neural Radiance Fields for Dynamic Scenes},
  author={Pumarola, Albert and Corona, Enric and Pons-Moll, Gerard and Moreno-Noguer, Francesc},
  journal={arXiv preprint arXiv:2011.13961},
  year={2020}
}