Virtual Scenario Simulation and Modeling Framework in Autonomous Driving Simulators

Recently, virtual environment-based techniques to train sensor-based autonomous driving models have been widely employed due their efficiency. However, a simulated environment is required be highly similar its real-world counterpart ensure the applicability of such actual vehicles. Though advances in hardware and three-dimensional graphics engine technology enabled creation realistic environments, myriad scenarios occurring real world can only up limited extent. In this study, scenario simulation modeling framework that simulates behavior objects may encountered while proposed address problem. This maximizes number scenarios, types, experience environment. Furthermore, simulator was implemented evaluate performance framework.