オンライン・オフライン適応ドメインランダム化のベンチマーク比較
Online vs. Offline Adaptive Domain Randomization Benchmark
シミュレータのパラメータを実データで自動調整する適応ドメインランダム化手法（SimOpt, BayRn, DROID, DROPO）を、オンラインとオフラインの設定で比較するオープンベンチマークを提案した。
著者: Gabriele Tiboni, Karol Arndt, Giuseppe Averta, Ville Kyrki, Tatiana Tommasi
分類: cs.RO, cs.LG
原文アブストラクト
Physics simulators have shown great promise for conveniently learning reinforcement learning policies in safe, unconstrained environments. However, transferring the acquired knowledge to the real world can be challenging due to the reality gap. To this end, several methods have been recently proposed to automatically tune simulator parameters with posterior distributions given real data, for use with domain randomization at training time. These approaches have been shown to work for various robotic tasks under different settings and assumptions. Nevertheless, existing literature lacks a thorough comparison of existing adaptive domain randomization methods with respect to transfer performance and real-data efficiency. In this work, we present an open benchmark for both offline and online methods (SimOpt, BayRn, DROID, DROPO), to shed light on which are most suitable for each setting and task at hand. We found that online methods are limited by the quality of the currently learned policy for the next iteration, while offline methods may sometimes fail when replaying trajectories in simulation with open-loop commands. The code used will be released at https://github.com/gabrieletiboni/adr-benchmark.
