Run the following files to generate sparse depth priors for all the three datasets for all the three input configurations.
cd src/prior_generators/visibility/
python VisibilityMask01_RealEstate.py
python VisibilityMask02_NeRF_LLFF.py
python VisibilityMask05_DTU.py
cd ../../../Running the above files creates a new directory data/databases/<DATABASE_NAME>/data/all/visibility_prior, which contains three sub-directories named VW02,VW03,VW04 corresponding to two, three and four input-view settings. Each of these directories will contain multiple sub-directories, one for every scene in the dataset. The following tree shows an exmaple.
data/databases/NeRF_LLFF/data/all/estimated_depths
|--VW02
| |--fern
| | |--visibility_masks
| | | |--0006.npy
| | | |--0006.png
| | | |--0013.npy
| | | |--0013.png
| | |--visibility_weights
| | | |--0006.npy
| | | |--0006.png
| | | |--0013.npy
| | | |--0013.png
| |--flower
| ...
|--VW03
| |--fern
| ...
|--VW04
| |--fern
| ...