Modified fusibile version for MVSNet post-processing, which produces similar point cloud results to paper:
MVSNet: Depth Inference for Unstructured Multi-view Stereo, ECCV 2018
The original source code is for the paper:
Massively Parallel Multiview Stereopsis by Surface Normal Diffusion, ICCV 2015
sudo apt-get install mesa-common-dev sudo apt-get install libgl1-mesa-dev libglu1-mesa-dev
cmake -DCUDA_TOOLKIT_ROOT_DIR=/usr/local/cuda-11.5/ -DCMAKE_C_COMPILER=gcc-10 -DCMAKE_CXX_COMPILER=g++-10 . make