Mapillary Vistas Dataset

A diverse street-level imagery dataset with pixel accurate and instance specific human annotations for understanding street scenes around the world.

Note: The dataset is free for research use.*

Data and Resources

Additional Info

Field Value
Source Mapillary
Maintainer Mapillary Vistas
Associated Tasks Object Recognition; Semantic Understanding
Geographical Area 6 Continents
Number of Instances 25,000 Images
Package Description Annotations: Banner, Barrier, Bench, Bicycle, Bicyclist, Bike Lane, Bike Rack, Billboard, Bird, Boat, Bridge, Building, Bus, CCTV Camera, Car, Car Mount, Caravan, Catch Basin, Crosswalk - Plain, Curb, Curb Cut, Desert, Dynamic, Ego Vehicle, Fence, Fire Hydrant, Garage, Ground, Ground Animal, Guard Rail, Junction Box, Lane Marking - Crosswalk, Lane Marking - Dashed Line, Lane Marking - Other, Lane Marking - Solid Line, Lane Marking - Stop Line, Lane Marking - Text, Lane Separator, Mailbox, Manhole, Motorcycle, Motorcyclist, Mountain, On Rails, Other Rider, Other Vehicle, Parking, Parking Meter, Pedestrian Area, Person, Person Group, Phone Booth, Pole, Pole Group, Pothole, Rail Track, Ramp, Road, Road Shoulder, Sand, Service Lane, Sidewalk, Sky, Snow, Static, Street Light, Terrain, Traffic Cone, Traffic Island, Traffic Light - Ambiguous, Traffic Light - Cyclists, Traffic Light - General (Horizontal/Back), Traffic Light - General (Horizontal/Front), Traffic Light - General (Horizontal/Side), Traffic Light - General (Upright/Back), Traffic Light - General (Upright/Front), Traffic Light - General (Upright/Side), Traffic Light - Other, Traffic Light - Pedestrians, Traffic Light - Temporary, Traffic Sign (Back), Traffic Sign (Front), Traffic Sign - Direction (Back), Traffic Sign - Direction (Front), Traffic Sign - Parking, Traffic Sign - Temporary (Back), Traffic Sign - Temporary (Front), Traffic Sign - Unknown, Traffic Sign Frame, Trailer, Trash Can, Truck, Tunnel, Unlabeled, Utility Pole, Vegetation, Vehicle Group, Wall, Water, Water Valve, Wheeled Slow, Wire Group.
Dataset has missing values False