DIFF-IPPO: Diffusion-Based Informative Path Planning with Open-Vocabulary Belief Maps

2026-06-15

AI Summary

This work addresses the challenge of low object detection probability in complex semantic and open-vocabulary environments by proposing an information-driven path planning approach that integrates an open-vocabulary belief map generator with a diffusion model-based planner. It introduces, for the first time, diffusion probabilistic models to global trajectory generation based on non-Gaussian, multimodal belief maps, enabling efficient batch conditional trajectory sampling to comprehensively cover high-belief regions. Experimental results demonstrate that the method achieves normalized detection scores of 81.49%–86.55% across diverse scenarios and accomplishes first-target discovery within 3.5 minutes in a five-agent cooperative search-and-rescue task, significantly enhancing both search efficiency and robustness.

Abstract

Exploration and object search require robots to perceive their environment, identify regions of interest, and plan trajectories that improve target-detection likelihood or maximize information gain. Many IPP methods, especially in continuous environmental monitoring, rely on Gaussian-process belief models, while object-search settings often produce complex, multimodal belief maps from semantic or open-vocabulary perception. Global trajectory generation directly conditioned on such non-Gaussian belief maps remains comparatively underexplored. Although diffusion-based planners offer strong capabilities for modeling such distributions, their use in informative path planning remains limited. In this work, we propose DIFF-IPPO, a pipeline that integrates an open-vocabulary belief map generator with a diffusion-based planner for global trajectory generation over belief maps. The method generates trajectories that concentrate sensor coverage over high-belief regions, achieving normalized detection scores between 81.49% and 86.55% across different dataset scenarios. We validate the system in a simulated search-and-rescue scenario where the planner searches candidate building regions to locate a burning building. In this setting, a team of five drones using batched belief-map-conditioned trajectory generation achieves first detections in 3.5 minutes.

Problem

• informative path planning
• open-vocabulary belief maps
• non-Gaussian belief maps
• global trajectory generation
• object search

Innovation

• diffusion-based planning
• informative path planning
• open-vocabulary perception
• belief maps
• trajectory generation