Reinforcement Learning for Active Perception in Autonomous Navigation

TL;DR

This paper presents a reinforcement learning approach for autonomous robots that actively control onboard cameras to improve navigation safety and exploration in unknown environments.

Contribution

It introduces an end-to-end RL framework combining motion planning and active perception, with a novel voxel-based information reward for better situational awareness.

Findings

Achieves safer flight compared to fixed-camera baselines.

Induces intrinsic exploratory behaviors in the robot.

Balances goal-directed motion with active sensing.

Abstract

This paper addresses the challenge of active perception within autonomous navigation in complex, unknown environments. Revisiting the foundational principles of active perception, we introduce an end-to-end reinforcement learning framework in which a robot must not only reach a goal while avoiding obstacles, but also actively control its onboard camera to enhance situational awareness. The policy receives observations comprising the robot state, the current depth frame, and a particularly local geometry representation built from a short history of depth readings. To couple collision-free motion planning with information-driven active camera control, we augment the navigation reward with a voxel-based information metric. This enables an aerial robot to learn a robust policy that balances goal-directed motion with exploratory sensing. Extensive evaluation demonstrates that our strategy achieves safer flight compared to using fixed, non-actuated camera baselines while also inducing intrinsic exploratory behaviors.