Seeing the Un-Scene: Learning Amodal Semantic Maps for Room Navigation

TL;DR

This paper presents a learning-based method for room navigation that predicts amodal semantic maps of unseen areas, enabling better generalization and navigation accuracy in novel environments.

Contribution

It introduces a model that predicts amodal semantic top-down maps for unseen regions, improving room navigation by leveraging architectural regularities.

Findings

Predicts belief maps of hidden regions beyond current view.

Generalizes well to new, unseen environments.

Reducing navigation to point navigation enhances performance.

Abstract

We introduce a learning-based approach for room navigation using semantic maps. Our proposed architecture learns to predict top-down belief maps of regions that lie beyond the agent's field of view while modeling architectural and stylistic regularities in houses. First, we train a model to generate amodal semantic top-down maps indicating beliefs of location, size, and shape of rooms by learning the underlying architectural patterns in houses. Next, we use these maps to predict a point that lies in the target room and train a policy to navigate to the point. We empirically demonstrate that by predicting semantic maps, the model learns common correlations found in houses and generalizes to novel environments. We also demonstrate that reducing the task of room navigation to point navigation improves the performance further.