データ拡張下での深層Q学習の安定化：ConvNetとVision Transformer
Stabilizing Deep Q-Learning with ConvNets and Vision Transformers under Data Augmentation
データ拡張を用いたオフポリシー強化学習の不安定性が高分散なQターゲットに起因することを特定し、それを安定化する簡便な手法を提案。画像ベースRLのサンプル効率と汎化性能をConvNetとViTの両方で向上させた。
著者: Nicklas Hansen, Hao Su, Xiaolong Wang
分類: cs.LG, cs.CV, cs.RO
原文アブストラクト
While agents trained by Reinforcement Learning (RL) can solve increasingly challenging tasks directly from visual observations, generalizing learned skills to novel environments remains very challenging. Extensive use of data augmentation is a promising technique for improving generalization in RL, but it is often found to decrease sample efficiency and can even lead to divergence. In this paper, we investigate causes of instability when using data augmentation in common off-policy RL algorithms. We identify two problems, both rooted in high-variance Q-targets. Based on our findings, we propose a simple yet effective technique for stabilizing this class of algorithms under augmentation. We perform extensive empirical evaluation of image-based RL using both ConvNets and Vision Transformers (ViT) on a family of benchmarks based on DeepMind Control Suite, as well as in robotic manipulation tasks. Our method greatly improves stability and sample efficiency of ConvNets under augmentation, and achieves generalization results competitive with state-of-the-art methods for image-based RL in environments with unseen visuals. We further show that our method scales to RL with ViT-based architectures, and that data augmentation may be especially important in this setting.
