On the webpage under point 7.2 you state that the Hausdorff distance of model vertices was used to define object symmetries. From this set a subjective selection took place, to arrive at a subset, without symmetries that might be resolved via textures.
As this doesn't make a lot of sense for the T-LESS dataset (mostly textureless), was the full set S′_M used? Alternatively, is the sourcecode for obtaining global object symmetries available within this (or another) repo?
On the webpage under point 7.2 you state that the Hausdorff distance of model vertices was used to define object symmetries. From this set a subjective selection took place, to arrive at a subset, without symmetries that might be resolved via textures.
As this doesn't make a lot of sense for the T-LESS dataset (mostly textureless), was the full set S′_M used? Alternatively, is the sourcecode for obtaining global object symmetries available within this (or another) repo?
Thanks