Cityscapes is a large-scale database which focuses on semantic understanding of urban street scenes. It provides semantic, instance-wise, and dense pixel annotations for 30 classes grouped into 8 categories (flat surfaces, humans, vehicles, constructions, objects, nature, sky, and void). The dataset consists of around 5000 fine annotated images and 20000 coarse annotated ones. Data was captured in 50 cities during several months, daytimes, and good weather conditions. It was originally recorded as video so the frames were manually selected to have the following features: large number of dynamic objects, varying scene layout, and varying background.
Source: A Review on Deep Learning Techniques Applied to Semantic Segmentation
Image Source: https://www.cityscapes-dataset.com/dataset-overview/
Variants: Semi-Supervised Semantic Segmentation on Cityscapes 6.25% labeled, Semi-Supervised Semantic Segmentation on Cityscapes 12.5% labeled, Cityscapes with extra (no coarse labels), Cityscapes with extra (no coarse), Cityscapes heterogeneous, Cityscapes 6.25% labeled, Cityscapes 5% labeled, Cityscapes 2% labeled, Cityscapes 128x128, Cityscapes 93 labeled, Cityscapes 10% labeled, Cityscapes, Cityscapes-5K 256x512, Cityscapes-25K 256x512, Cityscapes val, Cityscapes test, Cityscapes Photo-to-Labels, Cityscapes Labels-to-Photo, Cityscapes 50% labeled, Cityscapes 25% labeled, Cityscapes 12.5% labeled, Cityscapes 100 samples labeled
This dataset is used in 8 benchmarks:
Recent papers with results on this dataset: