Energy-based Compositional Diffusion Planning
For robotic planning, ECD provides a principled energy-based framework that improves compositional diffusion planners for long-horizon tasks, addressing a fundamental flaw in existing stitching approaches.
Energy-based Compositional Diffuser (ECD) formulates long-horizon robotic planning as minimizing the sum of local bridge potentials, correcting the non-conservative nature of heuristic stitching. It achieves state-of-the-art success rates on OGBench stitching tasks while maintaining inference speed comparable to heuristic methods.
Compositional diffusion planners aim to solve long-horizon robotic tasks using short training trajectories. Yet, current approaches often rely on the heuristic stitching of local predictions. We show that the resulting stitched update is generally a non-conservative field} that does not mathematically correspond to any valid global trajectory log-density function. We propose Energy-based Compositional Diffuser (ECD), a framework that formulates the global trajectory as the minimizer of the sum of local bridge potentials. This energy-based perspective defines a conservative correction field and contains a boundary reaction term that heuristic stitching omits. To enable efficient inference, we further introduce a Markov-based score approximation that computes the reaction term via a single block-tridiagonal solve, maintaining time complexity linear in the planning horizon. Empirically, ECD achieves state-of-the-art success rates on a range of OGBench stitching tasks, while nearly matching the inference speed of heuristic stitching methods. Code is available at https://github.com/GradientSpaces/ECD.