Skip to content
Inference code and trained networks for Deep Model-Based 6D Pose Refinement in RGB
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
config init with linemod benchvise model Jan 9, 2019
models init with linemod benchvise model Jan 9, 2019
refiner init with linemod benchvise model Jan 9, 2019
rendering init with linemod benchvise model Jan 9, 2019
utils init with linemod benchvise model Jan 9, 2019
.gitignore init with linemod benchvise model Jan 9, 2019
LICENSE Initial commit Jul 9, 2018
README.md Update README.md Jan 9, 2019
__init__.py init with linemod benchvise model Jan 9, 2019
test_refinement.py

README.md

This code accompanies the paper

Fabian Manhardt, Wadim Kehl, Nassir Navab and Federico Tombari: Deep Model-Based 6D Pose Refinement in RGB. ECCV 2018.(PDF)

and allows to reproduce parts of our work. Note that due to IP issues we can only provide our trained networks and the inference part. This allows to refine 6D hypotheses from monocular data.

Unfortunately, the code for training cannot be made available.

In order to use the code, you need to downloa the used datasets (hinterstoisser, tejani) in SIXD format (e.g. from here ) and use the test_refinement.py script to do the magic. Invoke 'python3 test_refinement.py --help' to see the available commands.

Note that we only provide the network for obj_02 of the LineMOD dataset. The other networks will be also released soon.

You can’t perform that action at this time.