Data Sets & Release 8 / by Jeffrey Johnson

Four new simulation data sets from Parallel Domain are now available, in addition to a new code release.

Data Sets:

These medium resolution data sets with semantic labels cover various scenarios with agent interaction and varying map topologies. Click the titles to go to the data set entry for each.

s02: Truck approach from behind

00206.jpg
  • 215 frame, medium resolution visual of an approach on a truck from behind on an uneven street
  • Ground truth, per-pixel semantic segmentation with blur noise
  • 1280x720 resolution
  • Includes source simulation movies
00023.jpg
  • 300 frame, medium resolution visual of an approach and traversal of a forked road
  • Ground truth, per-pixel semantic segmentation with blur noise
  • 1280x720 resolution
  • Includes source simulation movies
00057.jpg
  • 300 frame, medium resolution visual of an approach and traversal of a forked road
  • Ground truth, per-pixel semantic segmentation with blur noise
  • 1280x720 resolution
  • Includes source simulation movies
00055.jpg
  • 175 frame, medium resolution visual of traveling along a hilly road with transverse traffic
  • Ground truth, per-pixel semantic segmentation with blur noise
  • 1280x720 resolution
  • Includes source simulation movies

Updated Packages:

Release 8 is a minor code update that enables segmentation taxonomies to specify RGB labels with ranges in addition to specific values:

Development packages: https://is.gd/keniqu

Major changes and updates described below:

segmentation_taxonomy:

  • Taxonomy updated to describe RGB label ranges
  • Parsing rules updated to interpret RGB label ranges

segmentation_isp_field:

  • Data types and stream overloads updated