CycleIK

Overview

Implementation of CycleIK - Neuro-inspired Inverse Kinematics

CycleIK used in a human-in-the-loop grasping scenario on the Neuro-Inspired Collaborator (NICOL)

Neuro-inspired inverse kinematics training architecture. A batch of poses X is inferenced by either the MLP or GAN to obtain the batch of joint angles Θ. The joint angles are transformed back to Cartesian space X by the FK(Θ) function to be evaluated by the multi-objective function under constraints L.

Publications

Inverse Kinematics for Neuro-Robotic Grasping with Humanoid Embodied Agents. (2024). 
Habekost, JG., Gäde, C., Allgeuer, P., Wermter, S. 
Accepted at IROS 2024.

Arxiv: https://arxiv.org/html/2404.08825

CycleIK: Neuro-inspired Inverse Kinematics. (2023). 
Habekost, JG., Strahl, E., Allgeuer, P., Kerzel, M., Wermter, S. 
In: Artificial Neural Networks and Machine Learning – ICANN 2023. 
Lecture Notes in Computer Science, vol. 14254, pp. 457-470. 

Arxiv: https://arxiv.org/abs/2307.11554
Open Access: https://link.springer.com/chapter/10.1007/978-3-031-44207-0_38

Installation

Clone and install requirements

Best practice is to use anaconda or miniconda

git clone https://github.com/jangerritha/CycleIK.git
cd cycleik/
pip install -r requirements.txt
pip install -e .

Usage

from cycleik_pytorch import CycleIK

cycleik = CycleIK(robot="nicol", cuda_device='0', verbose=True, chain="right_arm")

start_pose = np.array([0.3, -0.5, 0.99, 0.0015305, 0.0009334, 0.70713, 0.70708])

target_pose = np.array([0.6, -0.3, 0.895, -0.5, 0.5, 0.5, 0.5])

control_points = np.array([np.array([0.3, -0.5, 1.2, -0.5, 0.5, 0.5, 0.5]),
                           np.array([0.6, -0.3, 1.2, -0.5, 0.5, 0.5, 0.5])])

bezier_curve = cycleik.generate_cubic_bezier_trajectory(start_pose=start_pose,
                                                        target_pose=target_pose,
                                                        control_points=control_points,
                                                        points=100)

joint_trajectory, _, _, _ = cycleik.inverse_kinematics(bezier_curve, calculate_error=True)

Test

The MLP network can be tested with the following command, remove the --cuda flag incase you are running a CPU-only system.

Exchange <robot> by one of the following options: nicol, nico, fetch, panda, or valkyrie.

The <chain> tag refers to the kinematic chain that is selected for evaluation.
Must match one of the move groups in the config yaml of the corresponding robot.

python test.py --cuda --robot <robot> --chain <chain> --network MLP 

Example

python test.py --cuda --robot nicol --chain right_arm --network MLP 

Download pretrained weights

Pre-trained weights are available for each of the five robot platforms on which CycleIK was evaluated.

cd <path_to_repo_location>
mkdir weights/<robot>
cd weights
mkdir <robot>

You can download a zip that contains all pre-trained weights from here.

Train

The training is executed very similar to running the tests.

python train.py --cuda --robot <robot> --chain <chain> --network MLP --epochs 10

Example

python train.py --cuda --robot nicol --chain right_arm --network MLP --epochs 10

Download dataset

For every robot there is a train, test, and validation dataset. Every dataset is delivered in a single file and must be added under <path_to_repo_location>/data.

cd <path_to_repo_location>
mkdir data

You can download a zip that contains all datasets used for our publication from here.

Contributing

If you find a bug, create a GitHub issue, or even better, submit a pull request. Similarly, if you have questions, simply post them as GitHub issues.