DR2L: ドメインランダム化強化学習でコーナーケースを顕在化し自動運転を堅牢化
DR2L: Surfacing Corner Cases to Robustify Autonomous Driving via Domain Randomization Reinforcement Learning
シミュレーション内の自動運転強化学習に敵対的モデルを導入し、ドメインランダム化で難しい事例を段階的に浮かび上がらせて実世界への転移を容易にする研究。
著者: Haoyi Niu, Jianming Hu, Zheyu Cui, Yi Zhang
分類: cs.RO, cs.AI, cs.LG
原文アブストラクト
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.
