SEA: Semantic Map Prediction for Active Exploration of Uncertain Areas

TL;DR

SEA introduces a semantic map prediction method combined with reinforcement learning to improve active robot exploration, enabling more efficient coverage of uncertain areas by predicting missing map regions and guiding exploration based on long-term understanding.

Contribution

The paper presents a novel iterative prediction-exploration framework with a reinforcement learning-based reward mechanism for long-term semantic map prediction and exploration.

Findings

Outperforms state-of-the-art exploration strategies.

Achieves higher map coverage within limited steps.

Effectively predicts missing map areas for better exploration.

Abstract

In this paper, we propose SEA, a novel approach for active robot exploration through semantic map prediction and a reinforcement learning-based hierarchical exploration policy. Unlike existing learning-based methods that rely on one-step waypoint prediction, our approach enhances the agent's long-term environmental understanding to facilitate more efficient exploration. We propose an iterative prediction-exploration framework that explicitly predicts the missing areas of the map based on current observations. The difference between the actual accumulated map and the predicted global map is then used to guide exploration. Additionally, we design a novel reward mechanism that leverages reinforcement learning to update the long-term exploration strategies, enabling us to construct an accurate semantic map within limited steps. Experimental results demonstrate that our method significantly outperforms state-of-the-art exploration strategies, achieving superior coverage ares of the global map within the same time constraints.