Integrating Deep RL and Bayesian Inference for ObjectNav in Mobile Robotics

arXiv cs.RO / 3/27/2026

💬 OpinionSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper tackles the object navigation/search problem in indoor mobile robotics by addressing partial observability, perceptual uncertainty, and the exploration–efficiency trade-off.
It proposes a hybrid framework that couples Bayesian inference—via an online spatial belief map updated from calibrated detections—with a deep reinforcement learning policy that selects navigation actions from that probabilistic state.
The Bayesian component explicitly represents uncertainty, while the RL component learns adaptive action selection without relying on handcrafted heuristics.
Experiments in realistic indoor simulation (Habitat 3.0) across two environments show improved success rates and reduced search effort versus baseline strategies.
Overall results indicate that combining probabilistic belief estimation with learned policies can yield more efficient and reliable object-search behavior under uncertainty.

Abstract

Autonomous object search is challenging for mobile robots operating in indoor environments due to partial observability, perceptual uncertainty, and the need to trade off exploration and navigation efficiency. Classical probabilistic approaches explicitly represent uncertainty but typically rely on handcrafted action-selection heuristics, while deep reinforcement learning enables adaptive policies but often suffers from slow convergence and limited interpretability. This paper proposes a hybrid object-search framework that integrates Bayesian inference with deep reinforcement learning. The method maintains a spatial belief map over target locations, updated online through Bayesian inference from calibrated object detections, and trains a reinforcement learning policy to select navigation actions directly from this probabilistic representation. The approach is evaluated in realistic indoor simulation using Habitat 3.0 and compared against developed baseline strategies. Across two indoor environments, the proposed method improves success rate while reducing search effort. Overall, the results support the value of combining Bayesian belief estimation with learned action selection to achieve more efficient and reliable objectsearch behavior under partial observability.