Every now and then a new dataset appears that aims to help training computer vision models for autonomous vehicles (AVs), being KITTI [1] the most famous, but others like Waymo [2] or nuScenes [3] can be found that try to innovate by adding more sensors, giving more data about the ego-vehicle surroundings or offering new scenarios. These approaches tend to produce a high quality dataset, but at the cost of time and money. The key difficulty is to create datasets with enough data to generalize multiple road scenarios and hazard use-cases in the training process of perception models for AVs. To address this concern, we propose AD PerDevKit, a perception development kit for autonomous driving based on CARLA simulator [4] able to generate an infinite set of annotated data referred to the ego-vehicle by recording virtual car sensors (camera, LiDAR, Radar) in virtual environments. Our ground-truth generator provides real-time information of the objects surrounding the vehicle that are visible for each sensor by publishing them on a ROS [5] topic. In addition, a dataset generated with this tool is also presented for some specific challenging scenarios and an evaluation of its contribution in the improvement of 2D object detection methods on real data is included. The code of our tool and its database is available in the following git repository: https://github.com/Javier-DlaP/ad_perdevkit.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    AD PerDevKit: An Autonomous Driving Perception Development Kit using CARLA simulator and ROS




    Publication date :

    2022-10-08


    Size :

    2497128 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Autonomous Navigation with Deep Reinforcement Learning in Carla Simulator

    Wang, Peilin / Technische Universität Dresden | SLUB | 2023


    A Waypoint Tracking Controller for Autonomous Vehicles Using CARLA Simulator

    Srinivas, Chinmay / Patil, Sharanbassappa S. | Springer Verlag | 2022


    Quantitative Evaluation of Autonomous Driving in CARLA

    Gao, Shang / Paulissen, Spencer / Coletti, Mark et al. | IEEE | 2021


    An Evaluation of "Crash Prediction Networks" (CPN) for Autonomous Driving Scenarios in CARLA Simulator

    Nair, Saasha / Shafaei, Sina / Auge, Daniel et al. | TIBKAT | 2021

    Free access

    Train Here, Drive There: Simulating Real-World Use Cases with Fully-Autonomous Driving Architecture in CARLA Simulator

    Gómez-Huélamo, Carlos / Del Egido, Javier / Bergasa, Luis M. et al. | Springer Verlag | 2020