Skip to end of metadata
Go to start of metadata

The zip files below contain 41 original KITTI images and the corresponding manually semantically labelled data. The labelled data is in the form of a Matlab .mat file with each entry in the array corresponding to the class label.

Example images and corresponding labelled images



Original KITTI ImagesMatlab .mat Files with Semantic Labels and KITTI_Images_Part


Each .mat file contains an array "S" and a 1x9 cell "names". The value of each element of S represents the class for that corresponding pixel, where the classes are given in "names". A value of 0 is represents an unknown class.

If you use this dataset, please cite the following paper:

B. Upcroft, C. McManus, W. Churchill, W. Maddern and P. Newman, Lighting Invariant Urban Street Classification, International Conference on Robotics and Automation, 2014


Please ask Ben Upcroft ( if you have any questions.



  • No labels