Repo containing everything relating to my Bachelor's Thesis.
Open the NRPClient_VirtualEmbodiment scene in /Assets/EmbodimentDiscrepancy/Scenes/
Attach the trackers to the body as they are on the avatar in the centre of the scene
Start the scene, and step into the T-posing avatar, and match his pose. Then press 'A' on the keyboard (this can be adapted easily in the /Assets/EmbodimentDiscrepancy/Scripts/TrackerAssignment/TrackedObjectAssigner.cs script).
The TrackedObjectAssigner will assign the correct trackers to the corresponding limbs.
The Neurorobotics_Unity3D_Client needs to be set up according to it's README. Then,in order to use it with the NRP, update the NRP as follows:
enter the docker image with:
docker exec -it nrp bash
tested with NRP in following state, might work with newer versions as well:
GazeboRosPackages on branch NRPRES-42 commit df0eb2a23634680f0deb76e0e81a66e272a1558b
Models on branch NRPRES-44 commit 4ece6f911081679407321ab8765061e528140d37
to see those branches, you might need to make sure all branches are fetched:
in .git/config change
fetch: refs/head/master
to
refs/head/*
after updating those modules:
in Models:
./create-symlinks.sh
in GazeboRosPackages:
catkin_make
don't forget to update Models in NRPClient:
Models repo should be at same commit as in NRP docker image
in project root dir:
python Assets/Models/scripts/copy_flat_hierarchy.py <PATH_TO_NRP_MODELS>