QuickNeRF

A reimplementation of Hash Encoding in InstantNGP for NeRF with a simplified intro to NeRF.

Hash encoding is one of the tricks used by InstantNGP to reduce the training time for NeRF to seconds instead of hours!! For more information and background, we highly recommend you to go through the references and watch our presentation.

This was a self-defined course project for Deep Learning course at KTH.

Experiment Table

Experiment with RTX 2060 Max-Q

	Model size	Step	PSNR	SSIM	LPIPS
LEGO baseline	1192k	50k	28.73	0.937	0.073
LEGO hash	19.2k	10k	28.45	0.940	0.068
MIC baseline	1192k	50k	30.53	0.967	0.051
MIC hash	19.2k	10k	28.19	0.960	0.055
SHIP baseline	1192k	50k	27.34	0.834	0.194
SHIP hash	19.2k	10k	26.37	0.834	0.175
(llff) ORCHID hash	19.2k	10k	17.58	0.688	0.227
(llff) ORCHID baseline	1192k	50k	20.57	0.686	0.297

T experiment

T	PSNR	SSIM	LPIPS
13	25.87	0.878	0.141
16	27.55	0.921	0.094
19	28.45	0.940	0.067
21	28.55	0.943	0.061

Citation

Kudos to the authors for their amazing results and code implementations:

@article{mueller2022instant,
    author = {Thomas M\"uller and Alex Evans and Christoph Schied and Alexander Keller},
    title = {Instant Neural Graphics Primitives with a Multiresolution Hash Encoding},
    journal = {ACM Trans. Graph.},
    issue_date = {July 2022},
    volume = {41},
    number = {4},
    month = jul,
    year = {2022},
    pages = {102:1--102:15},
    articleno = {102},
    numpages = {15},
    url = {https://doi.org/10.1145/3528223.3530127},
    doi = {10.1145/3528223.3530127},
    publisher = {ACM},
    address = {New York, NY, USA},
}

@misc{mildenhall2020nerf,
    title={NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis},
    author={Ben Mildenhall and Pratul P. Srinivasan and Matthew Tancik and Jonathan T. Barron and Ravi Ramamoorthi and Ren Ng},
    year={2020},
    eprint={2003.08934},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}

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