From the root path of the project:

docker build -f docker/Dockerfile -t neuralbody .

You may want to try several times since there are so many packages to be downloaded through the Internet and htpp(s) erros could occur.

The docker image contains the environment you need to run the project, while you still need to manually download data as described in INSTALL.md.

Note that the files downloaded are originally tar.gz files, while you need to extract each of them.

An example is like:

for name in $(ls *.tar.gz); do tar -xvf $name; done

Suppose you are at the root path of the project, run a docker container like:

docker run -it --rm --gpus=all \
--mount type=bind,source="$(pwd)",target=/app \
--mount type=bind,source=<DATAPATH>,target=/app/data \
neuralbody <COMMAND>

where <COMMAND> can be obtained from README.md and <DATAPATH> is your path for data.