DARLA: Improving Zero-shot Transfer In Reinforcement Learning

Abstract

Domain adaptation is an important open problem in deep reinforcement learning (RL). In many scenarios of interest data is hard to obtain, so agents may learn a source policy in a setting where data is readily available, with the hope that it generalises well to the target domain. We propose a new multi-stage RL agent, DARLA (DisentAngled Representation Learning Agent), which learns to see before learning to act. DARLA's vision is based on learning a disentangled representation of the observed environment. Once DARLA can see, it is able to acquire source policies that are robust to many domain shifts - even with no access to the target domain. DARLA significantly outperforms conventional baselines in zero-shot domain adaptation scenarios, an effect that holds across a variety of RL environments (Jaco arm, DeepMind Lab) and base RL algorithms (DQN, A3C and EC).

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyhiggins2017darla

Related papers

• On Zero-shot Reinforcement Learning (2025)0.00
• Adarl: What, Where, And How To Adapt In Transfer Reinforcement Learning (2021)0.00
• DRED: Zero-shot Transfer In Reinforcement Learning Via Data-regularised Environment Design (2024)1.81
• Off-dynamics Reinforcement Learning: Training For Transfer With Domain Classifiers (2020)0.00
• Zero-shot Generalization Of Vision-based RL Without Data Augmentation (2024)0.00
• Off-dynamics Reinforcement Learning Via Domain Adaptation And Reward Augmented Imitation (2024)0.00
• A Unified Framework For Zero-shot Reinforcement Learning (2025)0.00
• Inferring Behavior-specific Context Improves Zero-shot Generalization In Reinforcement Learning (2024)0.95