Jittor 可微渲染新视角生成比赛 JHashNeRF

Instant-NGP recently introduced a Multi-resolution Hash Encoding for neural graphics primitives like NeRFs. The original NVIDIA implementation mainly in C++/CUDA, based on tiny-cuda-nn, can train NeRFs upto 100x faster! This project is a pure Jittor implementation of Instant-NGP, built with the purpose of enabling AI Researchers to play around and innovate further upon this method. This project is built on top of the super-useful NeRF-pytorch, HashNeRF-pytorch, jrender implementation.

Description

This project contains the code implementation for the second Computational Graphics Challenge - Differentiable Rendering for Novel View Synthesis. As described above, the notable aspects of this project involve the utilization of Jittor to implement multi-resolution hashing encoding on the original NeRF (Neural Radiance Fields) model. It includes the addition of sparse loss and TV loss, as well as incorporating one-time importance sampling. The rendered images roughly resemble the following.

Uasge

Requirements

ubuntu 20.04 LTS
python >= 3.7
jittor >= 1.3.0

Please run the following script to create the environment

pip install -r requirements.txt

Dataset

bash download_competition_data.sh

Train

Run the following script to train

bash train.sh
# or
python run_nerf.py --config ./configs/Scar.txt

Eval

python val.py --config ./configs/$scene.txt --ft_path=lohs/$scene.tar

Acknowledgement

The project bases on jrender, HashNeRF-pytorch, NeRF-pytorch. Thanks to all of them.

@article{hu2020jittor,
  title={Jittor: a novel deep learning framework with meta-operators and unified graph execution},
  author={Hu, Shi-Min and Liang, Dun and Yang, Guo-Ye and Yang, Guo-Wei and Zhou, Wen-Yang},
  journal={Science China Information Sciences},
  volume={63},
  number={222103},
  pages={1--222103},
  year={2020}
}

@misc{lin2020nerfpytorch,
  title={NeRF-pytorch},
  author={Yen-Chen, Lin},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished={\url{https://github.com/yenchenlin/nerf-pytorch/}},
  year={2020}
}

@misc{bhalgat2022hashnerfpytorch,
  title={HashNeRF-pytorch},
  author={Yash Bhalgat},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished={\url{https://github.com/yashbhalgat/HashNeRF-pytorch/}},
  year={2022}
}

whao22/JHashNeRF