MSeg Documentation¶
MSeg is a framework for the ArmarX robot development environment, built to run and evaluate motion segmentation algorithms.
It provides the core functionality to easily integrate motion segmentation algorithms written in various programming languages. To integrate a motion segmentation algorithm, a supplied Programming Language Interface (PLI) can be used.
For more information about the structure of MSeg, please refer to the architectural overview.
Note
If you encounter a problem in the documentation or in MSeg, feel free to open a new issue in the issue tracker or send an email to incoming+h2t/kit-mseg/mseg@gitlab.com (this will create an issue in the issue tracker).
Installation and Setup¶
Guides on how to compile MSeg and the PLIs from source.
Integrate a Motion Segmentation Algorithm¶
Guides on how to integrate a motion segmentation algorithm into the core module.
- Integrate a C++ Algorithm
- Integrate a Java Algorithm
- Integrate a Python Algorithm
- Integrate a MATLAB Algorithm
If you are indecisive about the programming language, we would recommend to use Python.
Further Reference¶
Publication¶
- PDF version of the paper
- BibTeX:
@INPROCEEDINGS {Dreher2017,
author = {Christian R. G. Dreher and Nicklas Kulp and Christian Mandery and Mirko W\"achter and Tamim Asfour},
title = {A Framework for Evaluating Motion Segmentation Algorithms},
booktitle = {IEEE/RAS International Conference on Humanoid Robots (Humanoids)},
pages = {83--90},
year = {2017},
}