Awesome

Streamable Neural Fields

Paper link

Junwoo Cho*, Seungtae Nam*, Daniel Rho, Jong Hwan Ko, Eunbyung Park†<br> * Equal contribution, alphabetically ordered.<br> † Corresponding author.

European Conference on Computer Vision (ECCV), 2022

Overview

"Berliner Philharmoniker" © Stephan Rabold

0. Requirements

Setup a conda environment using commands below:

conda env create -f environment.yml
conda activate snf

1. Dataset

Download Kodak dataset from here.

Download UVG dataset from here.<br> When downloading UVG video, please use this version:<br>

• Resolution: 1080p<br>
• Bit depth: 8<br>
• Format: AVC<br>
• Container: MP4<br>

Download 3D point cloud dataset from here.

'data/' directory must be in your working directory. The structure is as follows:

Data layout

data/
    kodak/
        kodim01.png
        ...
        kodim24.png
    shape/
        armadillo.xyz
        dragon.xyz
        happy_buddha.xyz
    uvg/
        Beauty.mp4
        ...
        YachtRide.mp4

2. Reproducing experiments

Run the commands below.

Image spectral growing

bash scripts/train_image_spectral.sh

Image spatial growing

bash scripts/train_image_spatial.sh

Video temporal growing

bash scripts/train_video_temporal.sh

SDF spectral growing

bash scripts/train_sdf_spectral.sh

3. Results

You can find both qualitative and quantitative results in \results directory.

Citation

@inproceedings{cho2022streamable,
  title={Streamable neural fields},
  author={Cho, Junwoo and Nam, Seungtae and Rho, Daniel and Ko, Jong Hwan and Park, Eunbyung},
  booktitle={European Conference on Computer Vision},
  pages={595--612},
  year={2022},
  organization={Springer}
}