Organisation CoR-Lab -- Research Institute for Cognition and Robotics, Bielefeld University
Projects AMARSi
Project Descriptions This benchmark framework evaluates the performance of reaching motion generation approaches.The performance of these approaches is evaluated on a wide range of performance measures and typical tasks in robotics. The evaluation in this open source MATLAB software framework is done in the context of a default training data-set of human motions, which specifies the ground truth in the comparisons. The systematic comparisons allow to identify strengths and weaknesses of competing approaches and statistical evaluation of generalization and robustness. Robustness is tested by a chosen set of perturbation scenarios, which will interfere with the motion generation.
Title Amarsi Benchmark Framework
Version v1.0.0
Date date of the current version of the data in ISO EN 28601: 2014-08-28 16:26
Creators Andre Lemme, Mohammad Khansari-Zadeh
Contributors Yaron Meirovitch, Sebastien Gay
License Amarsi Benchmark Framework: Copyright(c) 2014 Andre Lemme, Mohammad Khansari-Zadeh, Yaron Meirovitch, Sebastien Gay. Find updates here . This software may be licensed under the terms of the GNU Lesser General Public License Version 3 (the ``LGPL''), or (at your option) any later version. Software distributed under the License is distributed on an ``AS IS'' basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See the LGPL for the specific language governing rights and limitations. You should have received a copy of the LGPL along with this program. If not, go to http://www.gnu.org/licenses/lgpl.html or write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
Keywords movement primitive, benchmark, comparisons, dynamical systems learning, movement generation
Format matlab files
Relation Lemme, Andre ; Meirovitch, Yaron ; Khansari-Zadeh, Seyed Mohammad ; Flash, Tamar ; Aude Billard ; Jochen J. Steil; Multi-criteria benchmarking of movement generating dynamical systems for learning-from-demonstrations
Acknowledgements The development of this software was funded by FP7 under GA. No. 248311-AMARSi.
Citation Information Paladyn, Journal of Behavioral Robotics. Volume 6, Issue 1, ISSN (Online) 2081-4836, DOI: 10.1515/pjbr-2015-0002, March 2015

toolkit.xml Magnifier (4.023 KB) Cord Wiljes, 2014-08-29 12:18