search

robot-learning-freiburg / MM-DistillNet

PyTorch code for training MM-DistillNet for multimodal knowledge distillation. http://rl.uni-freiburg.de/research/multimodal-distill
GNU General Public License v3.0
59 stars 14 forks source link

Clarification about depth images in the dataset #14

Open krantiparida opened 3 years ago

krantiparida commented 3 years ago

Hi, Thanks for the effort and for making the dataset public.

I have a small doubt about the provided depth images of the dataset. From the depth images, it seems cv2.COLORMAP_JET was used for converting the normalized depth values to the colormap image. But in the examples, it seems that the nearby points (points with lower depth values) are marked by red color whereas it should be blue as per the colormap. Can you please clarify this discrepancy or are you storing the inverse depth in the image?

drydenwiebe commented 2 years ago

The depth images seem to be created with https://github.com/pat676/MLPDepthMap/blob/master/src/MLPDataInterpreter.py#L115