```
@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={1–21},
year={2020}
}
@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},
}
@inproceedings{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},
booktitle={ECCV},
}
第二届计图(Jittor)人工智能挑战赛
赛题二:可微渲染新视角生成赛题
简介
本项目主要基于由 jittor 框架开发的 JNeRF 实现。
安装
本项目的运行环境要求
步骤1:安装依赖包
推荐使用 conda 创建环境
如果在安装过程中有任何问题,可以参考 Jittor.
步骤2:安装 JNeRF
安装完成之后可以在 Python 解释器中运行
import jnerf
测试 JNeRF是否安装成功。模型训练
本项目训练数据为 Jittor 大赛提供的合成数据,具体操作在 ./configs/*.py/ 中。
nerf_synthetic
,通过运行bash download_data.sh
即可获取。在训练时,必须修改配置文件中的数据路径configs
文件夹下,分别为Car.py, Coffee.py, Easyship.py, Scar.py, Scarf.py
。本项目训练用到的 GPU 是Tesla v100 32G
,单卡训练一个模型的时间大概是三个小时,训练结束会在./logs
文件夹下得到每个模型对用的test
渲染结果,在比赛A榜
和B榜
中得到的结果如下:A榜
B榜
测试
训练结束之后会在
./logs/<场景名>/
下自动保存参数文件*.pkl
. 也可以加载预训练模型进行测试和渲染,预训练参数模型可从此链接下载,下载完成后在根目录下新建./log
文件夹,并将参数文件组织成以下形式可以运行以下代码进行测试,只对
<文件名>
进行修改即可。运行结束会在
./result/<文件名>/test
文件夹下得到相应场景的测试视角图片。也可以运行以下代码,运行结束会
./result/<文件名>/
生成多视角展示的.mp4
格式的视频引用
``` @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={1–21}, year={2020} } @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}, } @inproceedings{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}, booktitle={ECCV}, }