DR2L: Surfacing Corner Cases to Robustify Autonomous Driving via Domain Randomization Reinforcement Learning

07/25/2021

∙

How to explore corner cases as efficiently and thoroughly as possible has long been one of the top concerns in the context of deep reinforcement learning (DeepRL) autonomous driving. Training with simulated data is less costly and dangerous than utilizing real-world data, but the inconsistency of parameter distribution and the incorrect system modeling in simulators always lead to an inevitable Sim2real gap, which probably accounts for the underperformance in novel, anomalous and risky cases that simulators can hardly generate. Domain Randomization(DR) is a methodology that can bridge this gap with little or no real-world data. Consequently, in this research, an adversarial model is put forward to robustify DeepRL-based autonomous vehicles trained in simulation to gradually surfacing harder events, so that the models could readily transfer to the real world.

READ FULL TEXT

DR2L: Surfacing Corner Cases to Robustify Autonomous Driving via Domain Randomization Reinforcement Learning

From Simulation to Real World Maneuver Execution using Deep Reinforcement Learning

Enhanced Transfer Learning for Autonomous Driving with Systematic Accident Simulation

(Re)^2H2O: Autonomous Driving Scenario Generation via Reversely Regularized Hybrid Offline-and-Online Reinforcement Learning

Cycle-Consistent World Models for Domain Independent Latent Imagination

Sim-To-Real Transfer for Miniature Autonomous Car Racing

Sim-to-Real via Sim-to-Sim: Data-efficient Robotic Grasping via Randomized-to-Canonical Adaptation Networks

Sim-to-Real Transfer in Deep Reinforcement Learning for Robotics: a Survey

DR2L: Surfacing Corner Cases to Robustify Autonomous Driving via Domain Randomization Reinforcement Learning

Related Research

From Simulation to Real World Maneuver Execution using Deep Reinforcement Learning

Enhanced Transfer Learning for Autonomous Driving with Systematic Accident Simulation

(Re)^2H2O: Autonomous Driving Scenario Generation via Reversely Regularized Hybrid Offline-and-Online Reinforcement Learning

Cycle-Consistent World Models for Domain Independent Latent Imagination

Sim-To-Real Transfer for Miniature Autonomous Car Racing

Sim-to-Real via Sim-to-Sim: Data-efficient Robotic Grasping via Randomized-to-Canonical Adaptation Networks

Sim-to-Real Transfer in Deep Reinforcement Learning for Robotics: a Survey