Awesome
<p align="center"> <h2 align="center"><strong>LoopGaussian: Creating 3D Cinemagraph with Multi-view Images via Eulerian Motion Field</strong></h2> <p align="center"> <span> Jiyang Li<sup>1*</sup>, <a href="https://scholar.google.com/citations?user=PKFAv-cAAAAJ&hl=en">Lechao Cheng</a><sup>2*</sup>, Zhangye Wang<sup>1</sup>, Tingting Mu<sup>3</sup>, Jingxuan He<sup>2ā</sup> </span> <br> <span> <sup>*</sup>Equal contribution. <sup>ā</sup>Corresponding author. <br> <sup>1</sup>Zhejiang University, <sup>2</sup>Hefei University of Technology, <sup>3</sup>The University of Manchester </span> </p> <div align="center"><a href='https://arxiv.org/abs/2404.08966'><img src='https://img.shields.io/badge/arXiv-2404.08966-b31b1b.svg'></a> <a href='https://pokerlishao.github.io/LoopGaussian/'><img src='https://img.shields.io/badge/Project-Page-Green'></a>
</div> <!-- ## News -->Environment
- CUDA = 12.1
- Pytorch = 2.1.1
- Python = 3.11
pip install -r requirements.txt
Train
Train 3D-GS
python gaussian_splatting/train.py -s {images_path}
Train LoopGaussian
python train_loop.py --task_name ${task_name}
The rendering result will be saved in output/{task_name}/visual.mp4