Code to process lidar information into semantic segmentation.

TODO list in the near future:

Data processing side:

• Properly transform from lidar intensities to colors to allow a better visualization.
• Get a correspondence between lidar points and camera pixels.

Deep Learning side:

• Code all the DL model! D:
• Test it.
• Fix it.
• Jump to three tasks ago.