NagabhushanSN95 / ViP-NeRF

Official code release accompanying the paper - "ViP-NeRF: Visibility Prior for Sparse Input Neural Radiance Fields"
MIT License
68 stars 7 forks source link

Confused of Extract Dataset #4

Closed kkennethwu closed 1 year ago

kkennethwu commented 1 year ago

Hi, thanks for your great work. I have a problem about extract dataset of RealEstate10K. I ran python VideoNameMapper.py, python SceneSelector01.py successfully. And when I ran python DataExtractor01.py, I got this error:

截圖 2023-08-11 下午2 18 48

And I'm confused of what these three python file do. Below is the config in SceneSelector01.py and DataExtractor.py :

截圖 2023-08-11 下午2 24 24 截圖 2023-08-11 下午2 25 58
kkennethwu commented 1 year ago

Also, if I modify the range to 10 in data_extractor.extract_data(feature, range(500)), I got this error.

截圖 2023-08-11 下午2 39 38
NagabhushanSN95 commented 1 year ago

Hi, thanks for your interest in our work. The data extraction for RealEstate-10K is not configured properly yet. I will update this thread once that is done. It may take another month or so. However, the data extraction pipeline for the other datasets should work.

If you want to experiment with RealEstate-10K dataset, you can follow the guidelines on the RealEstate-10K dataset webpage to download it. For any other datasets, you can follow a similar approach. Then place the files similar to how the LLFF/DTU datasets are placed and you should be good to go!

NagabhushanSN95 commented 1 year ago

Please check this documentation for an example of how to place the data.

Or, you can place the data however you prefer and modify the data loader appropriately

kkennethwu commented 1 year ago

I see ! big thanks for your help. I've tried nerf-llff successfully.

NagabhushanSN95 commented 1 year ago

Closing this issue.