Our CVPR 2016 paper presenting the Cityscapes Dataset is now available. We include a detailed analysis of the dataset, baseline results, and discussions.
Our evaluation server is now online, where you can upload your results on the Cityscapes test set, they get evaluated and you receive an email once this is done.
Examples of our coarse annotations can be found here. These annotations are intended to support research that aims to exploit large volumes of (weakly) annotated data, e.g. for training deep neural networks.