CARLA: An Open Urban Driving Simulator

• DOI: 10.48550/arxiv.1711.03938
• Publication Date: 2017-01-01
• Authors: Alexey Dosovitskiy; Germán Ros; Felipe Codevilla; Antonio M. López; Vladlen Koltun
• Keywords: Driving simulator; Open platform

We introduce CARLA, an open-source simulator for autonomous driving research. CARLA has been developed from the ground up to support development, training, and validation of urban systems. In addition code protocols, provides open digital assets (urban layouts, buildings, vehicles) that were created this purpose can be used freely. The simulation platform supports flexible specification sensor suites environmental conditions. use study performance three approaches driving: a classic modular pipeline, end-to-end model trained via imitation learning, reinforcement learning. are evaluated in controlled scenarios increasing difficulty, their is examined metrics provided by illustrating platform's utility supplementary video viewed at https://youtu.be/Hp8Dz-Zek2E