Generalization to Novel Objects using Prior Relational Knowledge
This addresses the challenge of sample-efficient generalization for AI agents in tasks with unseen objects, though it is incremental in combining existing techniques.
The paper tackles the problem of enabling agents to generalize to novel objects in new environments by reasoning over prior relational knowledge, achieving 5-10x greater sample efficiency than baselines in Sokoban and Pacman environments.
To solve tasks in new environments involving objects unseen during training, agents must reason over prior information about those objects and their relations. We introduce the Prior Knowledge Graph network, an architecture for combining prior information, structured as a knowledge graph, with a symbolic parsing of the visual scene, and demonstrate that this approach is able to apply learned relations to novel objects whereas the baseline algorithms fail. Ablation experiments show that the agents ground the knowledge graph relations to semantically-relevant behaviors. In both a Sokoban game and the more complex Pacman environment, our network is also more sample efficient than the baselines, reaching the same performance in 5-10x fewer episodes. Once the agents are trained with our approach, we can manipulate agent behavior by modifying the knowledge graph in semantically meaningful ways. These results suggest that our network provides a framework for agents to reason over structured knowledge graphs while still leveraging gradient based learning approaches.