Learning Interactive World Model for Object-Centric Reinforcement Learning
This work addresses the need for more robust and transferable policies in reinforcement learning by explicitly modeling object interactions, though it is incremental as it builds on existing object-centric methods.
The paper tackled the problem of object-centric reinforcement learning by introducing the Factored Interactive Object-Centric World Model (FIOC-WM), which learns structured representations of objects and their interactions, resulting in improved sample efficiency and generalization for policy learning on simulated robotic and embodied-AI benchmarks.
Agents that understand objects and their interactions can learn policies that are more robust and transferable. However, most object-centric RL methods factor state by individual objects while leaving interactions implicit. We introduce the Factored Interactive Object-Centric World Model (FIOC-WM), a unified framework that learns structured representations of both objects and their interactions within a world model. FIOC-WM captures environment dynamics with disentangled and modular representations of object interactions, improving sample efficiency and generalization for policy learning. Concretely, FIOC-WM first learns object-centric latents and an interaction structure directly from pixels, leveraging pre-trained vision encoders. The learned world model then decomposes tasks into composable interaction primitives, and a hierarchical policy is trained on top: a high level selects the type and order of interactions, while a low level executes them. On simulated robotic and embodied-AI benchmarks, FIOC-WM improves policy-learning sample efficiency and generalization over world-model baselines, indicating that explicit, modular interaction learning is crucial for robust control.