Skip to content

This project explores Robotic Path Planning Using Diffusion Models (Janner et al., 2023) in 2D and 3D environments. The project was completed for Advanced Deep Learning in Robotics (ADLR) course at TU Munich at the Learning AI for Dexterous Robots lab of Prof. Dr. Berthold Bäuml, in cooperation with the German Aerospace Center (DLR).

Notifications You must be signed in to change notification settings

adlr-team/diffusion-path-planning

Repository files navigation

Robotic Path Planning Using Diffusion Models

This project was conducted as part of the Advanced Deep Learning in Robotics (ADLR) course at the Learning AI for Dexterous Robots lab of Prof. Dr. Berthold Bäuml, in cooperation with the German Aerospace Center (DLR). The aim was to explore the usage of diffusion models for robotic path planning. We reference the work of Janner et al. Diffuser.

We explored diffusion models for path planning in 2D and 3D environments. Initially applied to the 2D Pointmaze-medium, we then extended our work to the Kuka Robot LWR3.

2D Pointmaze-Medium

2D Pointmaze-medium 2D Pointmaze-medium

Kuka Robot LWR3

Kuka Robot LWR3 Diffused Q-Values

About

This project explores Robotic Path Planning Using Diffusion Models (Janner et al., 2023) in 2D and 3D environments. The project was completed for Advanced Deep Learning in Robotics (ADLR) course at TU Munich at the Learning AI for Dexterous Robots lab of Prof. Dr. Berthold Bäuml, in cooperation with the German Aerospace Center (DLR).

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •