長期的タスクに取り組む身体性エージェントのためのシーン記憶トランスフォーマー
Scene Memory Transformer for Embodied Agents in Long-Horizon Tasks
観測を記憶に蓄積しアテンションで時空間依存性を活用する記憶ベース方策SMTを提案し、視覚ナビゲーションタスクで既存手法を大きく上回る性能を示した。
著者: Kuan Fang, Alexander Toshev, Li Fei-Fei, Silvio Savarese
分類: cs.LG, cs.CV, cs.RO, stat.ML
原文アブストラクト
Many robotic applications require the agent to perform long-horizon tasks in partially observable environments. In such applications, decision making at any step can depend on observations received far in the past. Hence, being able to properly memorize and utilize the long-term history is crucial. In this work, we propose a novel memory-based policy, named Scene Memory Transformer (SMT). The proposed policy embeds and adds each observation to a memory and uses the attention mechanism to exploit spatio-temporal dependencies. This model is generic and can be efficiently trained with reinforcement learning over long episodes. On a range of visual navigation tasks, SMT demonstrates superior performance to existing reactive and memory-based policies by a margin.
