« View All Resources
Semantic Labels for a subset of KITTI
- Posted on February 13, 2015 at 3:54 am.
We have manually labelled 146 images from the KITTI dataset, at pixel level, with 10 semantic categories (Sky, Building, Road, Sidewal, Fence, Vegetation, Pole, Car, Sign, Pedestrian). Our data is currently being used for training and evaluation of image parsing algorithms in urban enviroments. The original RGB images are taken from the Visual Odometry database (KITTI