Learning to Map for Active Semantic Goal Navigation
This addresses the problem of navigating to find objects in unknown scenes for robotics or AI agents, representing an incremental advance over existing methods.
The paper tackles object goal navigation in unseen indoor environments by proposing a framework that actively learns to generate semantic maps beyond the agent's field of view and uses uncertainty to set long-term goals, showing improved results on the Matterport3D dataset over competitive baselines.
We consider the problem of object goal navigation in unseen environments. Solving this problem requires learning of contextual semantic priors, a challenging endeavour given the spatial and semantic variability of indoor environments. Current methods learn to implicitly encode these priors through goal-oriented navigation policy functions operating on spatial representations that are limited to the agent's observable areas. In this work, we propose a novel framework that actively learns to generate semantic maps outside the field of view of the agent and leverages the uncertainty over the semantic classes in the unobserved areas to decide on long term goals. We demonstrate that through this spatial prediction strategy, we are able to learn semantic priors in scenes that can be leveraged in unknown environments. Additionally, we show how different objectives can be defined by balancing exploration with exploitation during searching for semantic targets. Our method is validated in the visually realistic environments of the Matterport3D dataset and show improved results on object goal navigation over competitive baselines.