Bayesian Generalized Kernel Inference for Exploration of Autonomous Robots

TL;DR

This paper introduces a Bayesian generalized kernel inference method for robot exploration that predicts mutual information and confidence efficiently, enabling scalable decision-making in complex environments.

Contribution

It develops a lightweight, continuous inference model for mutual information prediction that reduces computational complexity and balances local and global exploration strategies.

Findings

Achieves logarithmic complexity in decision-making

Maintains exploration performance in complex scenes

Provides open-source implementation for the robotics community

Abstract

This paper concerns realizing highly efficient information-theoretic robot exploration with desired performance in complex scenes. We build a continuous lightweight inference model to predict the mutual information (MI) and the associated prediction confidence of the robot's candidate actions which have not been evaluated explicitly. This allows the decision-making stage in robot exploration to run with a logarithmic complexity approximately, this will also benefit online exploration in large unstructured, and cluttered places that need more spatial samples to assess and decide. We also develop an objective function to balance the local optimal action with the highest MI value and the global choice with high prediction variance. Extensive numerical and dataset simulations show the desired efficiency of our proposed method without losing exploration performance in different environments. We also provide our open-source implementation codes released on GitHub for the robot community.