This repository provides the evaluation code of the publication
"AIMY: An Open-source Table Tennis Ball Launcher for Versatile and High-fidelity Trajectory Generation"
accepted to ICRA 2023.
It contains scripts for recording, preprocessing and visualisation of evaluation data. Furthermore, we provide the suite for target shooting with a feed-forward neural network.
AIMY is a custom-designed three-wheeled table tennis ball launcher designed for advancing table tennis robot research.
Basic dependencies:
pip install numpy, pandas, scipy, matplotlib, seaborn, pytest, h5py, tensorflow
Misc dependencies:
pip install isort, black, flake8, mypy, tickzplotlib, sphinx
Interfacing MPI Intelligent Soft Robots Lab libraries:
- ball_launcher beepy: low-level control software of the table tennis ball launcher
- tennicam_client: ball position tracking system
- signal_handler: tracks interrupts via terminal
The low-level control software for AIMY can be found here.
To cite AIMY in your academic research, please use the following bibtex lines:
@misc{dittrich2023aimy,
title={AIMY: An Open-source Table Tennis Ball Launcher for Versatile and High-fidelity Trajectory Generation},
author={Alexander Dittrich and Jan Schneider and Simon Guist and Nico Gürtler and Heiko Ott and Thomas Steinbrenner and Bernhard Schölkopf and Dieter Büchler},
year={2023},
eprint={2210.06048},
archivePrefix={arXiv}
}BSD 3-Clause license.
Copyright © 2023, Max Planck Gesellschaft