Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces the Graph-Guided Safe Diffuser (G2SD), a hierarchical framework that combines a topological graph planner with a diffusion model to enhance safety in trajectory planning. By abstracting the data manifold into a learned latent graph, G2SD effectively guides the diffusion process, mitigating the risk of kinematic infeasibility due to manifold rupture. Experimental results show that G2SD significantly increases the goal-reaching rate without collisions in Maze2D navigation, achieving rates of 98% compared to 40-50% with traditional methods.
G2SD boosts goal-reaching success from 40-50% to an impressive 98% in collision-free navigation tasks by leveraging topological graph guidance.
Many diffusion-based planners enforce safety through inference-time guidance, but such interleaved trajectory deformations often degrade kinematic feasibility due to manifold rupture. We propose Graph-Guided Safe Diffuser (G2SD), a hierarchical framework that leverages a high-level topological graph planner to guide a low-level diffusion model. G2SD enforces safety at a structural level by abstracting the data manifold into a learned latent graph, on which high-level planning is performed. Continuous trajectories are generated by diffusion planners, which are conditioned on the graph node representations selected by the high-level planner. Theoretical analyses demonstrate conditions under which manifold rupture occurs in diffusion planners, and show that G2SD improves safety by reducing the constraint violation probability as the number of segments increases. Experiments demonstrate that G2SD substantially outperforms baselines, increasing goal-reaching rate without any collision from 40-50% to 98% in Maze2D navigation and also achieving superior task scores in locomotion.