Statements (57)
| Predicate | Object |
|---|---|
| gptkbp:instance_of |
gptkb:Database_Management_System
|
| gptkbp:bfsLayer |
4
|
| gptkbp:bfsParent |
gptkb:Mask_R-CNN
gptkb:Seg_Net |
| gptkbp:analysis |
visualization tools available
|
| gptkbp:application |
computer vision, machine learning
|
| gptkbp:available_formats |
images, annotations
|
| gptkbp:category |
road, vehicles, pedestrians
|
| gptkbp:challenges |
real-world application of AI
|
| gptkbp:class |
gptkb:30
|
| gptkbp:collaborations |
partnerships with tech companies
various universities and research institutions |
| gptkbp:collection |
driving through cities
|
| gptkbp:contains |
street scenes
|
| gptkbp:contribution |
open-source projects
|
| gptkbp:created_by |
gptkb:Max_Planck_Institute_for_Informatics
|
| gptkbp:data_privacy |
considerations for data use
|
| gptkbp:data_type |
JPEG images
benchmark dataset manual and semi-automated supports various ML frameworks |
| gptkbp:data_usage |
advancing urban scene understanding
limited to urban scenes normalization, augmentation over 25 GB |
| gptkbp:features |
high-resolution images
|
| gptkbp:focus |
urban environments
|
| gptkbp:focus_area |
German cities
|
| gptkbp:genetic_diversity |
various weather conditions
|
| gptkbp:has_artwork |
5000
|
| https://www.w3.org/2000/01/rdf-schema#label |
Cityscapes dataset
|
| gptkbp:image |
camera footage
|
| gptkbp:is_cited_in |
cited in numerous papers
|
| gptkbp:is_divided_into |
train, val, test
|
| gptkbp:is_documented_in |
gptkb:JSON
pixel-level annotations |
| gptkbp:is_evaluated_by |
mean Intersection over Union (m Io U)
|
| gptkbp:is_popular_in |
widely used in research
|
| gptkbp:is_used_for |
semantic segmentation
|
| gptkbp:label |
multi-class labeling
|
| gptkbp:license |
gptkb:Creative_Commons_Attribution_4.0_International_License
|
| gptkbp:performance |
CVPR, ECCV
|
| gptkbp:population_trend |
growing interest in urban datasets
|
| gptkbp:primary_source |
autonomous driving research
|
| gptkbp:products |
high-quality annotations
|
| gptkbp:provides_information_on |
encourages collaboration
free for research purposes image segmentation challenges |
| gptkbp:related_works |
AD E20 K
|
| gptkbp:release_year |
gptkb:2016
|
| gptkbp:resolution |
2048x1024
|
| gptkbp:scientific_goals |
improve scene understanding
|
| gptkbp:training |
Deep Lab, FCN
|
| gptkbp:tutorials |
Label Me, VGG Image Annotator
|
| gptkbp:updates |
periodic updates
|
| gptkbp:user_base |
researchers, developers, students
|
| gptkbp:website |
https://www.cityscapes-dataset.com
|