Mixed Density Diffuser: Efficient Planning with Non-uniform Temporal Resolution

2025-10-27

AI Summary

Traditional trajectory planning struggles to balance efficiency and accuracy under non-uniform temporal scales. To address this, this paper proposes a diffusion-based trajectory planning method with adaptive time-resolution control. The core contribution is the first introduction of a tunable hybrid time-density mechanism, which dynamically modulates planning density at critical time intervals via hyperparameters—thereby breaking the constraints of uniform-step paradigms and enhancing modeling of long-horizon temporal dependencies. The method integrates sparse-step training with a time-density-aware diffusion process optimization. Evaluated on D4RL benchmarks—including Maze2D, Franka Kitchen, and Antmaze—our approach achieves state-of-the-art (SOTA) performance in both planning efficiency and downstream policy execution quality.

Technology Category

• Planning, Routing, and Scheduling: Temporal Planning
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Multiagent Systems: Multiagent Planning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Web Mining and Content Analysis: Content-based information diffusion

Abstract

Recent studies demonstrate that diffusion planners benefit from sparse-step planning over single-step planning. Training models to skip steps in their trajectories helps capture long-term dependencies without additional or memory computational cost. However, predicting excessively sparse plans degrades performance. We hypothesize this temporal density threshold is non-uniform across a temporal horizon and that certain parts of a planned trajectory should be more densely planned. We propose Mixed Density Diffuser (MDD), a diffusion planner where the densities throughout the horizon are tunable hyperparameters. MDD achieves a new SOTA across the Maze2D, Franka Kitchen, and Antmaze D4RL task domains.

Problem

• Optimizing non-uniform temporal resolution in diffusion planning
• Addressing performance degradation from excessive step skipping
• Enabling tunable planning density across temporal horizons

Innovation

• Mixed Density Diffuser enables tunable temporal resolution planning
• Non-uniform density allocation optimizes sparse-step trajectories
• Hyperparameter-controlled diffusion achieves state-of-the-art performance