Occupancy Anticipation for Efficient Exploration and Navigation
This work addresses efficient exploration and navigation in 3D environments for robotics and AI agents, representing a novel method for a known bottleneck.
The paper tackles the problem of limited spatial awareness in navigation by proposing occupancy anticipation, where an agent uses RGB-D observations to infer occupancy beyond visible regions, resulting in significantly better performance than baselines and outperforming state-of-the-art methods on Gibson and Matterport3D datasets, including winning the 2020 Habitat PointNav Challenge.
State-of-the-art navigation methods leverage a spatial memory to generalize to new environments, but their occupancy maps are limited to capturing the geometric structures directly observed by the agent. We propose occupancy anticipation, where the agent uses its egocentric RGB-D observations to infer the occupancy state beyond the visible regions. In doing so, the agent builds its spatial awareness more rapidly, which facilitates efficient exploration and navigation in 3D environments. By exploiting context in both the egocentric views and top-down maps our model successfully anticipates a broader map of the environment, with performance significantly better than strong baselines. Furthermore, when deployed for the sequential decision-making tasks of exploration and navigation, our model outperforms state-of-the-art methods on the Gibson and Matterport3D datasets. Our approach is the winning entry in the 2020 Habitat PointNav Challenge. Project page: http://vision.cs.utexas.edu/projects/occupancy_anticipation/