This is really nice work.
In apps/volumetric_fusion/__main__.py, output_root_directory is defined as that where the output mesh is saved. This is incorrect and should be the directory containing RGBD frames + instrinsics.txt. For every video then, the output mesh gets saved in a sub-directory.
So basically, we only need to define 2 inputs per input video: (1) RGBD frames and (2) .pkl tracking file.
This is really nice work.
In
apps/volumetric_fusion/__main__.py,output_root_directoryis defined as that where the output mesh is saved. This is incorrect and should be the directory containing RGBD frames + instrinsics.txt. For every video then, the output mesh gets saved in a sub-directory.So basically, we only need to define 2 inputs per input video: (1) RGBD frames and (2)
.pkltracking file.