Switching Attention in Time-Varying Environments via Bayesian Inference of Abstractions

TL;DR

This paper presents a Bayesian inference-based method for robots to detect context changes in dynamic environments and switch attention mechanisms accordingly, improving focus on relevant information for reliable operation.

Contribution

It introduces a novel approach combining bisimulation-based state abstractions with Bayesian inference to detect context switches and adapt attention in real-time.

Findings

Successfully detects context changes online

Robustly ignores distractors in complex environments

Demonstrated on both discrete and continuous tracking tasks

Abstract

Motivated by the goal of endowing robots with a means for focusing attention in order to operate reliably in complex, uncertain, and time-varying environments, we consider how a robot can (i) determine which portions of its environment to pay attention to at any given point in time, (ii) infer changes in context (e.g., task or environment dynamics), and (iii) switch its attention accordingly. In this work, we tackle these questions by modeling context switches in a time-varying Markov decision process (MDP) framework. We utilize the theory of bisimulation-based state abstractions in order to synthesize mechanisms for paying attention to context-relevant information. We then present an algorithm based on Bayesian inference for detecting changes in the robot's context (task or environment dynamics) as it operates online, and use this to trigger switches between different abstraction-based attention mechanisms. Our approach is demonstrated on two examples: (i) an illustrative discrete-state tracking problem, and (ii) a continuous-state tracking problem implemented on a quadrupedal hardware platform. These examples demonstrate the ability of our approach to detect context switches online and robustly ignore task-irrelevant distractors by paying attention to context-relevant information.