Semi-parametric Topological Memory for Navigation
This addresses the problem of efficient navigation for AI agents in novel settings, representing a novel method for a known bottleneck rather than a foundational advance.
The paper tackles navigation in unseen environments by introducing a semi-parametric topological memory (SPTM) architecture, which combines a non-parametric graph and a parametric deep network to build topological maps from short footage, resulting in a threefold higher success rate in goal-directed navigation compared to the best baseline.
We introduce a new memory architecture for navigation in previously unseen environments, inspired by landmark-based navigation in animals. The proposed semi-parametric topological memory (SPTM) consists of a (non-parametric) graph with nodes corresponding to locations in the environment and a (parametric) deep network capable of retrieving nodes from the graph based on observations. The graph stores no metric information, only connectivity of locations corresponding to the nodes. We use SPTM as a planning module in a navigation system. Given only 5 minutes of footage of a previously unseen maze, an SPTM-based navigation agent can build a topological map of the environment and use it to confidently navigate towards goals. The average success rate of the SPTM agent in goal-directed navigation across test environments is higher than the best-performing baseline by a factor of three. A video of the agent is available at https://youtu.be/vRF7f4lhswo