I've been running EfficientPose on my own datasets, and I have been trying both gray and rgb datasets separately to compare results. I have noticed however that I should not be including the term 'rgb' in my dataset name...
[linemod.py:353] mask_paths = [img_path.replace("rgb", "mask") for img_path in image_paths]
The path name replaces 'rgb' into 'mask'. It took me a couple hours to find why it could not load my mask files.
Perhaps it is a good idea to only replace the last occurrence of rgb, so not to cause errors in the wording of a dataset.
Hi there!
I've been running EfficientPose on my own datasets, and I have been trying both gray and rgb datasets separately to compare results. I have noticed however that I should not be including the term 'rgb' in my dataset name...
The path name replaces 'rgb' into 'mask'. It took me a couple hours to find why it could not load my mask files.
Perhaps it is a good idea to only replace the last occurrence of rgb, so not to cause errors in the wording of a dataset.