Cityscapes Dataset

The Cityscapes Dataset focuses on semantic understanding of urban street scenes. In the following, we give an overview on the design choices that were made to target the dataset’s focus.

Features

Polygonal annotations

• Dense semantic segmentation
• Instance segmentation for vehicle and people

Complexity

• 30 classes
• See Class Definitions for a list of all classes and have a look at the applied labeling policy.

Diversity

• 50 cities
• Several months (spring, summer, fall)
• Daytime
• Good/medium weather conditions
• Manually selected frames
• Large number of dynamic objects
• Varying scene layout
• Varying background

Volume

• 5 000 annotated images with fine annotations (examples)
• 20 000 annotated images with coarse annotations (examples)

Metadata

• Preceding and trailing video frames. Each annotated image is the 20^th image from a 30 frame video snippets (1.8s)
• Corresponding right stereo views
• GPS coordinates
• Ego-motion data from vehicle odometry
• Outside temperature from vehicle sensor

Extensions by other researchers

• Bounding box annotations of people
• Images augmented with fog and rain

Benchmark suite and evaluation server

• Pixel-level semantic labeling
• Instance-level semantic labeling
• Panoptic semantic labeling 

Labeling Policy

Labeled foreground objects must never have holes, i.e. if there is some background visible ‘through’ some foreground object, it is considered to be part of the foreground. This also applies to regions that are highly mixed with two or more classes: they are labeled with the foreground class. Examples: tree leaves in front of house or sky (everything tree), transparent car windows (everything car).