Below, you can find all currently available downloads. A README and various scripts for inspection, preparation, and evaluation can be found in our git repository.
Download
The following packages are available for download. Please download only those packages that you need for your immediate research.
gtFine_trainvaltest.zip (241MB) [md5] fine annotations for train and val sets (3475 annotated images) and dummy annotations (ignore regions) for the test set (1525 images)
gtCoarse.zip (1.3GB) [md5] coarse annotations for train and val set (3475 annotated images) and train_extra (19998 annotated images)
leftImg8bit_trainvaltest.zip (11GB) [md5] left 8-bit images - train, val, and test sets (5000 images)
leftImg8bit_trainextra.zip (44GB) [md5] left 8-bit images - trainextra set (19998 images, note that the image "troisdorf_000000_000073_leftImg8bit.png" is corrupt/black)
rightImg8bit_trainvaltest.zip (11GB) [md5] right 8-bit images - train, val, and test sets (5000 images)
rightImg8bit_trainextra.zip (44GB) [md5] right 8-bit images - trainextra set (19998 images)
disparity_trainvaltest.zip (3.5GB) [md5] precomputed depth maps using SGM for train, val, and test sets (5000 images)
disparity_trainextra.zip (15GB) [md5] precomputed depth maps using SGM for trainextra set
camera_trainvaltest.zip (2MB) [md5] intrinsic and extrinsic camera parameters for train, val, and test sets
camera_trainextra.zip (8MB) [md5] intrinsic and extrinsic camera parameters for trainextra set
vehicle_trainvaltest.zip (2MB) [md5] vehicle odometry + GPS coordinates + temperature for train, val, and test sets
vehicle_trainextra.zip (7MB) [md5] vehicle odometry + GPS coordinates + temperature for trainextra set
leftImg8bit_demoVideo.zip (6.6GB) [md5] video sequences for qualitative evaluation, left 8-bit images only
all_demoVideo.zip (59GB) [md5] video sequences for qualitative evaluation, all types of data
leftImg8bit_blurred.zip (11GB) [md5] Left 8-bit images with blurred faces and license plates. Please compute results on the original images but use the blurred ones for visualization. We thank Mapillary for blurring the images.
leftImg8bit_sequence_trainvaltest.zip (324GB) [md5] 30-frame snippets (17Hz) surrounding each left 8-bit image (-19 | +10) from the train, val, and test sets (150000 images)
timestamp_sequence.zip (40MB) [md5] timestamps for 30-frame snippets from train, val, and test sets
vehicle_sequence.zip (56MB) [md5] vehicle odometry + GPS coordinates + temperature for 30-frame snippets from train, val, and test sets
rightImg8bit_sequence_trainvaltest.zip (322GB) [md5] 30-frame snippets (17Hz) surrounding each right 8-bit image (-19 | +10) from the train, val, and test sets (150000 images)
disparity_sequence_trainvaltest.zip (106GB) [md5] precomputed depth maps using SGM for 30-frame snippets from train, val, and test sets
leftImg8bit_allFrames_frankfurt.zip (241GB) [md5] left 8-bit images for whole drive in frankfurt from val set
timestamp_allFrames_frankfurt.zip (29MB) [md5] timestamps for whole drive in frankfurt from val set
vehicle_allFrames_frankfurt.zip (38MB) [md5] vehicle odometry + GPS coordinates + temperature for whole drive in frankfurt from val set
leftImg16bit_trainvaltest.zip (37GB) [md5] left HDR images stored as 16-bit PNGs - train, val, and test sets (5000 images)
rightImg8bit_allFrames_frankfurt.zip (242GB) [md5] right 8-bit images for whole drive in frankfurt from val set
leftImg16bit_trainextra.zip (149GB) [md5] left HDR images stored as 16-bit PNGs - trainextra set (19998 images)
rightImg16bit_trainvaltest.zip (37GB) [md5] right HDR images stored as 16-bit PNGs - train, val, and test sets (5000 images)
rightImg16bit_trainextra.zip (149GB) [md5] right HDR images stored as 16-bit PNGs - trainextra set (19998 images)
The following packages from third party publications are available for download.
If you use the packages in your research, please cite both publications, Cityscapes and the work listed in conjunction with the package. Please contact the respective authors directly in case of questions.
gtBbox3d_trainvaltest.zip (6.6MB) [md5] 3D bounding box annotations of vehicles for train and val sets (3475 annotated images). Citation: N. Gählert, N. Jourdan, M. Cordts, U. Franke, and J. Denzler, "Cityscapes 3D: Dataset and Benchmark for 9 DoF Vehicle Detection", CVPRW 2020.
gtBbox_cityPersons_trainval.zip (2.2MB) [md5] CityPersons bounding box annotations of people for train and val sets (3475 annotated images). Citation: S. Zhang, R. Benenson, and B. Schiele, "CityPersons: A Diverse Dataset for Pedestrian Detection", CVPR 2017.
gtFinePanopticParts_trainval.zip (474MB) [md5] Panoptic parts annotations for train and val sets (3475 annotated images). Citation: P. Meletis, X. Wen, C. Lu, D. de Geus, and G. Dubbelman, "Cityscapes-Panoptic-Parts and PASCAL-Panoptic-Parts datasets for Scene Understanding", arXiv 2020.
leftImg8bit_trainval_foggyDBF.zip (20GB) [md5] Foggy DBF images (left 8-bit) for train and val sets (3475 images). Citation: C. Sakaridis, D. Dai, S. Hecker, and L. Van Gool, "Model Adaptation with Synthetic and Real Data for Semantic Dense Foggy Scene Understanding", ECCV 2018.
leftImg8bit_trainvaltest_foggy.zip (30GB) [md5] Foggy images (left 8-bit) for train, val, and test sets (5000 images). Citation: C. Sakaridis, D. Dai, and L. Van Gool, "Semantic Foggy Scene Understanding with Synthetic Data", IJCV 2018.
leftImg8bit_trainextra_foggy.zip (40GB) [md5] Foggy images (left 8-bit) for trainextra set (19997 images). Citation: C. Sakaridis, D. Dai, and L. Van Gool, "Semantic Foggy Scene Understanding with Synthetic Data", IJCV 2018.
leftImg8bit_trainval_rain.zip (21GB) [md5] Rain images (left 8-bit) for parts of train and val sets (295 images, 36 variants each). Citation: X. Hu, C.-W. Fu, L. Zhu, and P.-A. Heng, "Depth-Attentional Features for Single-Image Rain Removal", CVPR 2019.