The Neural Automation Project
The Neural Automation Project

BibTeX Collection

Klaus Neumann, collection of bibTeX entries
bibTeX entry collection for LaTeX.
NeumannCitations.txt
Text-Dokument [4.8 KB]

Thesis

Klaus Neumann, "Reliability of Extreme Learning Machines". Bielefeld University, 2014.
The reliable application of machine learning methods becomes increasingly important in challenging engineering domains. In particular, the application of extreme learning machines (ELM) seems promising because of their capability of very efficient processing of large and high-dimensional data sets. However, the ELM paradigm is based on the concept of randomly initialized and fixed input weights and is thus inherently unreliable. This black-box character usually repels engineers from application in potentially safety critical tasks. The goal of this thesis is therefore to equip the ELM approach with the abilities to perform in a reliable manner.
PhDThesis_kneumann_130214.pdf
PDF-Dokument [15.1 MB]
Klaus Neumann, "Reliability of Extreme Learning Machines". Bielefeld University, 2014.
bibTeX entry for LaTeX.
NeumannPhDThesis2014.txt
Text-Dokument [181 Bytes]

Journal Papers

K. Neumann and J. J. Steil, "Learning Robot Motions with Stable Dynamical Systems under Diffeomorphic Transformations". Robotics and Autonomous Systems, vol. 70(C), pp. 1-15, 2015.
Accuracy and stability have in recent studies been emphasized as the two major ingredients to learn robot motions from demonstrations with dynamical systems. Several approaches yield stable dynamical systems but are also limited to specific dynamics that can potentially result in a poor reproduction performance. The current work addresses this accuracy-stability dilemma through a new diffeomorphic transformation approach that serves as a framework generalizing the class of demonstrations that are learnable by means of provably stable dynamical systems. We apply the proposed framework to extend the application domain of the stable estimator of dynamical systems (SEDS) by generalizing the class of learnable demonstrations by means of diffeomorphic transformations tau . The resulting approach
NeumannSteil_RaAS2015.pdf
PDF-Dokument [8.3 MB]
K. Neumann and J. J. Steil, "Learning Robot Motions with Stable Dynamical Systems under Diffeomorphic Transformations". Robotics and Autonomous Systems, vol. 70(C), pp. 1-15, 2015.
bibTeX entry for LaTeX.
Neumann2015.txt
Text-Dokument [292 Bytes]
M. Rolf, K. Neumann, J. Queißer, F. Reinhart, A. Nordmann, J. J. Steil, "A Multi-Level Control Architecture for the Bionic Handling Assistant". Advanced Robotics vol. 29(13: SI), pp. 847-859, 2015.
The Bionic Handling Assistant is one of the largest soft continuum robots and very special in being a pneumatically operated platform that is able to bend, stretch, and grasp in all directions. It nevertheless shares many challenges with smaller continuum and other softs robots such as parallel actuation, complex movement dynamics, slow pneumatic actuation, non-stationary behavior, and a lack of analytic models. To master the control of this challenging robot, we argue for a tight integration of standard analytic tools, simulation, control, and state of the art machine learning into an overall architecture that can serve as blueprint for control design also beyond the BHA. To this aim, we show how to integrate specific modes of operation and different levels of control in a synergistic man
RolfNeumannQueisserReinhartNordmannSteil[...]
PDF-Dokument [3.0 MB]
M. Rolf, K. Neumann, J. Queißer, F. Reinhart, A. Nordmann, J. J. Steil, "A Multi-Level Control Architecture for the Bionic Handling Assistant". Advanced Robotics vol. 29(13: SI), pp. 847-859, 2015.
bibTeX entry for LaTeX.
Rolf2015.txt
Text-Dokument [378 Bytes]
A. Lemme, K. Neumann, R. F. Reinhart, J. J. Steil, "Neural Learning of Vector Fields for Encoding Stable Dynamical Systems". Neurocomputing (Special Issue ESANN), vol. 141, pp. 3-14, 2014.
This paper investigates the results of the previous paper in more detail in a special issue. In particular, the method is tested in a humanoid robot scenario involving iCub. The paper is published in the special issue of the ESANN 2013 conference.
LemmeNeumannReinhartSteil_NeuCom2014.pdf
PDF-Dokument [6.5 MB]
A. Lemme, K. Neumann, R. F. Reinhart, J. J. Steil, "Neural Learning of Vector Fields for Encoding Stable Dynamical Systems". Neurocomputing (Special Issue ESANN), vol. 141, pp. 3-14, 2014.
bibTeX entry for LaTeX.
Lemme2013b.txt
Text-Dokument [264 Bytes]
K. Neumann and J. J. Steil, "Optimizing Extreme Learning Machines via Ridge Regression and Batch Intrinsic Plasticity". Neurocomputing (Special Issue ELM 2012), vol. 102, pp. 23-30, 2013.
This paper analyzes batch intrinsic plasticity in more detail. In particular together with ridge regression for robotic applications. I presented the paper at the ELM 2011 in Hangzhou, China.
NeumannSteil_NeuCom2013.pdf
PDF-Dokument [1.5 MB]
K. Neumann and J. J. Steil, "Optimizing Extreme Learning Machines via Ridge Regression and Batch Intrinsic Plasticity". Neurocomputing (Special Issue ELM 2012), vol. 102, pp. 23-30, 2013.
bibTeX entry for LaTeX.
Neumann2013.txt
Text-Dokument [249 Bytes]
K. Neumann, C. Strub, and J. J. Steil, "Intrinsic Plasticity via Natural Gradient Descent with Application to Drift Compensation". Neurocomputing (Special Issue ESANN 2012), vol. 112, pp. 26-33, 2013
This paper extends the natural gradient descent for intrinsic plasticity by another modification of intrinsic plasticity and applies this novel learning rule to compensate input drifts. Main parts of this work considering the application to drift compensation and the working point transformation originate from the master thesis of Claudius Strub under my supervision.
NeumannStrubSteil_NeuCom2013.pdf
PDF-Dokument [2.1 MB]
K. Neumann, C. Strub, and J. J. Steil, "Intrinsic Plasticity via Natural Gradient Descent with Application to Drift Compensation". Neurocomputing (Special Issue ESANN 2012), vol. 112, pp. 26-33, 2013
bibTeX entry for LaTeX.
Neumann2013b.txt
Text-Dokument [261 Bytes]
K. Neumann, M. Rolf, and J. J. Steil, "Reliable Integration of Continuous Constraints into Extreme Learning Machines". Uncertainty, Fuzziness & Knowledge-Based Sys., vol. 21(2): pp. 35-50, 2013.
This paper proposes the ideas to use continuous constraints as prior knowledge and introduces an algorithm for implementation into extreme learning machines. It is shown that this algorithm also works in a real world scenario involving the bionic handling assistant. I presented the paper at the ELM 2013 in Singapore and received a best presentation award.
NeumannRolfSteil_IJUFKS2013.pdf
PDF-Dokument [1.2 MB]
K. Neumann, M. Rolf, and J. J. Steil, "Reliable Integration of Continuous Constraints into Extreme Learning Machines". Uncertainty, Fuzziness & Knowledge-Based Sys., vol. 21(2): pp. 35-50, 2013.
bibTeX entry for LaTeX.
Neumann2013c.txt
Text-Dokument [315 Bytes]
K. Neumann, C. Emmerich, and J. J. Steil, "Regularization by Intrinsic Plasticity and its Synergies with Recurrence for Random Projection Methods". JILSA, vol. 4, no. 3, pp. 230-246, 2012.
The Paper investigates the role of intrinsic plasticity as a feature regularization and recurrence as a technique to produce a non-linear mixture of sigmoid features for random projections. The paper was written during my PhD time but most of the actual content were developed in my master's thesis.
NeumannEmmerichSteil_JILSA2012.pdf
PDF-Dokument [1.4 MB]
K. Neumann, C. Emmerich, and J. J. Steil, "Regularization by Intrinsic Plasticity and its Synergies with Recurrence for Random Projection Methods". JILSA, vol. 4, no. 3, pp. 230-246, 2012.
bibTeX entry for LaTeX.
Neumann2012.txt
Text-Dokument [344 Bytes]

Conference Proceedings

J. Queißer, K. Neumann, M. Rolf, F. Reinhart, J.J. Steil, "An Active Compliant Control Mode for Interaction with a Pneumatic Soft Robot". IEEE/RSJ IROS, pp. 573-579, 2014.
Bionic soft robots offer exciting perspectives for more flexible and safe physical interaction with the world and humans. Unfortunately, their hardware design often prevents analytical modeling, which in turn is a prerequisite to apply classical automatic control approaches. On the other hand, also modeling by means of learning is hardly feasible due to many degrees of freedom, high-dimensional state spaces and the
softness properties like e.g. mechanical elasticity, which cause limited repeatability and complex dynamics. Nevertheless, the realization of basic control modes is important to leverage the potential of soft robots for applications. We therefore propose a hybrid approach combining classical and learning elements for the realization of an interactive control mode for an elastic
QueisserNeumannRolfReinhartSteil_IROS201[...]
PDF-Dokument [3.2 MB]
J. Queißer, K. Neumann, M. Rolf, F. Reinhart, J.J. Steil, "An Active Compliant Control Mode for Interaction with a Pneumatic Soft Robot". IEEE/RSJ IROS, pp. 573-579, 2014.
bibTeX entry for LaTeX.
Queisser2014.txt
Text-Dokument [325 Bytes]
A. Unger, W. Sextro, S. Althoff, K. Neumann, R. F. Reinhart, M. Broekelmann, D. Bolowski, K. Guth. "Investigation and modeling of the ultrasonic softening effect for the copper wire bonding process".
Proc. of Int Conf. on Integrated Power Electronics Systems, pp. 25-27, 2014.
This conference contribution investigates the ultrasonic softening effect for copper wire bonding and proposes an approach which models this effect with additonal use of prior domain knowledge about the task. The Incub project and partners provided the data set for data-driven learning and investigations about the ultrasonic softening effect for copper wire bonding. The paper will appear at the beginning of 2014 and Unger and I will present the paper at CIPS 2014 in Nurnberg, Germany.
UngerSextroAlthoffMeyerNeumannReinhartBr[...]
PDF-Dokument [726.2 KB]
A. Unger, W. Sextro, S. Althoff, K. Neumann, R. F. Reinhart, M. Broekelmann, D. Bolowski, K. Guth. "Investigation and modeling of the ultrasonic softening effect for the copper wire bonding process".
bibTeX entry for LaTeX.
Unger2014.txt
Text-Dokument [374 Bytes]
A. Lemme, K. Neumann, R. F. Reinhart, and J. J. Steil, "Neurally Imprinted Stable Vector Fields". Proc. Europ. Symp. on Artificial Neural Networks, pp. 327-332, 2013.
This paper introduces a neural network strategy to model dynamical systems learned from data, e.g. in the eld of imitation learning in robotics. The paper is a joint work of the European project "AMARSi" and the leading-edge cluster "it's owl". Andre Lemme and Felix Reinhart mainly developed the idea of movement generation through a vector eld representation while I developed the algorithm based on Lyapunov's stability theory to enforce asymptotic stability by means of the constrained learning scheme introduced during my PhD studies. I presented the paper at the ESANN 2013 in Bruges, Belgium and received a best student paper award.
LemmeNeumannReinhartSteil_ESANN2013.pdf
PDF-Dokument [1.9 MB]
A. Lemme, K. Neumann, R. F. Reinhart, and J. J. Steil, "Neurally Imprinted Stable Vector Fields". Proc. Europ. Symp. on Artificial Neural Networks, pp. 327-332, 2013.
bibTeX entry for LaTeX.
Lemme2013b.txt
Text-Dokument [264 Bytes]
K. Neumann, A. Lemme, and J. J. Steil, "Neural Learning of Stable Dynamical Systems based on Data-Driven Lyapunov Candidates". IEEE/RSJ IROS, pp. 1216-1222, 2013.
The paper proposes a method to learn Lyapunov candidate functions from data which I developed for the estimation of vector fields. I presented the paper at IROS 2013 in Tokyo, Japan.
NeumannLemmeSteil_IROS2013.pdf
PDF-Dokument [1.9 MB]
K. Neumann, A. Lemme, and J. J. Steil, "Neural Learning of Stable Dynamical Systems based on Data-Driven Lyapunov Candidates". IEEE/RSJ IROS, pp. 1216-1222, 2013.
bibTeX entry for LaTeX.
Neumann2013d.txt
Text-Dokument [305 Bytes]
K. Neumann and J. J. Steil, "Intrinsic Plasticity via Natural Gradient Descent". Proc. Europ. Symp. on Artificial Neural Networks, pp. 555-560, 2012.
This paper introduces the natural gradient descent for intrinsic plasticity which is interpreted as stochastic gradient descent. I presented the paper at ESANN 2012 in Bruges, Belgium.
NeumannSteil_ESANN2012.pdf
PDF-Dokument [317.6 KB]
K. Neumann and J. J. Steil, "Intrinsic Plasticity via Natural Gradient Descent". Proc. Europ. Symp. on Artificial Neural Networks, pp. 555-560, 2012.
bibTeX entry for LaTeX.
Neumann2012b.txt
Text-Dokument [244 Bytes]
K. Neumann and J. J. Steil, "Batch Intrinsic Plasticity for Extreme Learning Machines". Proc. Int. Conf. on Artificial Neural Networks, vol. 6791, no. 1, pp. 339-346, 2011.
This paper introduces batch intrinsic plasticity as a method to optimize extreme learning machines. The pretraining aims at desired output distributions of the hidden neurons and was inspired by intrinsic plasticity - a method for pretraining recurrent reservoir networks. I presented the paper at the ICANN 2011 in Helsinki, Finland.
NeumannSteil_ICANN2011.pdf
PDF-Dokument [549.5 KB]
K. Neumann and J. J. Steil, "Batch Intrinsic Plasticity for Extreme Learning Machines". Proc. Int. Conf. on Artificial Neural Networks, vol. 6791, no. 1, pp. 339-346, 2011.
bibTeX entry for LaTeX.
Neumann2011.txt
Text-Dokument [255 Bytes]
K. Neumann, M. Rolf, J. J. Steil, and M. Gienger, "Learning Inverse Kinematics for Pose-Constraint Bi-Manual Movements". Proc. of Int. Conf. on Simulation of Adaptive Behavior, 2010.
We present a neural network approach to learn inverse kinematics of the humanoid robot ASIMO, where we focus on bi-manual tool use. The learning copes with both the highly redundant inverse kinematics of ASIMO and the additional arbitrary constraint imposed by the tool that couples both hands. We show that this complex kinematics can be learned from few ground-truth examples using an efficient recurrent reservoir framework, which has been introduced previously for kinematics learning and movement generation. We analyze and quantify the network’s generalization for a given tool by means of reproducing the constraint in untrained target motions.
NeumannRolfSteil_SAB2010.pdf
PDF-Dokument [595.0 KB]
K. Neumann, M. Rolf, J. J. Steil, and M. Gienger, "Learning Inverse Kinematics for Pose-Constraint Bi-Manual Movements". Proc. of Int. Conf. on Simulation of Adaptive Behavior, 2010.
bibTeX entry for LaTeX.
Neumann2010.txt
Text-Dokument [317 Bytes]

Drafts

K. Neumann, J.J. Steil, "Building Neural Representations for Flexible Tool-Use". unpublished draft, 2013.
A novel approach to model tool-use for humanoid robots such as iCub is introduced. The learning represents the highly redundant coupling of the inverse kinematics and the motions imposed by the tool in a flexible way. It is shown that the manipulation kinematics can be learned from few ground-truth examples using an ecient extreme learning machine framework. The experiments reveal that this approach exhibits
human-like behavior (motor hysteresis eect) without learning from human data.
NeumannSteil_FlexibleToolUse2013.pdf
PDF-Dokument [1.4 MB]
K. Neumann, J.J. Steil, "Building Neural Representations for Flexible Tool-Use". unpublished draft, 2013.
bibTeX Entry for LaTeX.
Neumann2019.txt
Text-Dokument [245 Bytes]
Druckversion | Sitemap
© Klaus Neumann