ApolloScape Datasets

ApolloScape, part of the Apollo project for autonomous driving, is a research-oriented project to foster innovations in all aspects of autonomous driving, from perception, navigation, to control. It hosts open access to semantically annotated (pixel-level) street view images and simulation tools that supports user-defined polices. It is an evolving project that new datasets and new capability will be added regularly. Various workshops and challenges will be hosted to encourage exchanges of ideas and jointly advance the state of the art in autonomous driving research. ApolloScape contains large and rich labelling including holistic semantic dense point cloud for each site, stereo, per-pixel semantic labelling, lanemark labelling, instance segmentation, 3D car instance, high accurate location for every frame in various driving videos from multiple sites, cities and daytimes.

Data and Resources

Additional Info

Field Value
Source ApolloScape
Maintainer ApolloScape
Associated Tasks Image Parsing; Detection; Tracking; Localization; Segmentation
Number of Instances Evolving
Package Description Categories (class): Others (others, rover), Sky (sky), Movable object (car, car_groups, motorbicycle, motorbicycle_group, person, person_group, rider, rider_group, truck, truck_group, bus, bus_group, tricycle, tricycle_group) flat (road, siderwalk), Road obstacles (traffic_cone, road_pile, fence), Roadside objects (traffic_light), void (pole, traffic_sign, wall, dustbin, billboard), Building (building, bridge, tunnel, overpass), Natural (vegetation), Unlabeled (unlabeled).
Dataset has missing values False