GenPlanner: From Noise to Plans -- Emergent Reasoning in Flow Matching and Diffusion Models

TL;DR

GenPlanner leverages diffusion and flow matching models to generate paths in complex environments, demonstrating superior performance over traditional CNN methods through iterative trajectory refinement.

Contribution

The paper introduces GenPlanner, a novel generative approach for path planning using diffusion models and flow matching, with two variants: DiffPlanner and FlowPlanner.

Findings

FlowPlanner performs well with few generation steps.

The approach outperforms baseline CNN models.

Generative models effectively solve maze path planning.

Abstract

Path planning in complex environments is one of the key problems of artificial intelligence because it requires simultaneous understanding of the geometry of space and the global structure of the problem. In this paper, we explore the potential of using generative models as planning and reasoning mechanisms. We propose GenPlanner, an approach based on diffusion models and flow matching, along with two variants: DiffPlanner and FlowPlanner. We demonstrate the application of generative models to find and generate correct paths in mazes. A multi-channel condition describing the structure of the environment, including an obstacle map and information about the starting and destination points, is used to condition trajectory generation. Unlike standard methods, our models generate trajectories iteratively, starting with random noise and gradually transforming it into a correct solution. Experiments conducted show that the proposed approach significantly outperforms the baseline CNN model. In particular, FlowPlanner demonstrates high performance even with a limited number of generation steps.