Header logo is


2017


Thumb xl amd intentiongan
Multi-Modal Imitation Learning from Unstructured Demonstrations using Generative Adversarial Nets

Hausman, K., Chebotar, Y., Schaal, S., Sukhatme, G., Lim, J.

In Proceedings from the conference "Neural Information Processing Systems 2017., (Editors: Guyon I. and Luxburg U.v. and Bengio S. and Wallach H. and Fergus R. and Vishwanathan S. and Garnett R.), Curran Associates, Inc., Advances in Neural Information Processing Systems 30 (NIPS), December 2017 (inproceedings)

am

pdf video [BibTex]

2017


pdf video [BibTex]


Thumb xl fig toyex lqr1kernel 1
On the Design of LQR Kernels for Efficient Controller Learning

Marco, A., Hennig, P., Schaal, S., Trimpe, S.

Proceedings of the 56th IEEE Annual Conference on Decision and Control (CDC), pages: 5193-5200, IEEE, IEEE Conference on Decision and Control, December 2017 (conference)

Abstract
Finding optimal feedback controllers for nonlinear dynamic systems from data is hard. Recently, Bayesian optimization (BO) has been proposed as a powerful framework for direct controller tuning from experimental trials. For selecting the next query point and finding the global optimum, BO relies on a probabilistic description of the latent objective function, typically a Gaussian process (GP). As is shown herein, GPs with a common kernel choice can, however, lead to poor learning outcomes on standard quadratic control problems. For a first-order system, we construct two kernels that specifically leverage the structure of the well-known Linear Quadratic Regulator (LQR), yet retain the flexibility of Bayesian nonparametric learning. Simulations of uncertain linear and nonlinear systems demonstrate that the LQR kernels yield superior learning performance.

am ics pn

arXiv PDF On the Design of LQR Kernels for Efficient Controller Learning - CDC presentation DOI Project Page [BibTex]

arXiv PDF On the Design of LQR Kernels for Efficient Controller Learning - CDC presentation DOI Project Page [BibTex]


Thumb xl robot legos
Interactive Perception: Leveraging Action in Perception and Perception in Action

Bohg, J., Hausman, K., Sankaran, B., Brock, O., Kragic, D., Schaal, S., Sukhatme, G.

IEEE Transactions on Robotics, 33, pages: 1273-1291, December 2017 (article)

Abstract
Recent approaches in robotics follow the insight that perception is facilitated by interactivity with the environment. These approaches are subsumed under the term of Interactive Perception (IP). We argue that IP provides the following benefits: (i) any type of forceful interaction with the environment creates a new type of informative sensory signal that would otherwise not be present and (ii) any prior knowledge about the nature of the interaction supports the interpretation of the signal. This is facilitated by knowledge of the regularity in the combined space of sensory information and action parameters. The goal of this survey is to postulate this as a principle and collect evidence in support by analyzing and categorizing existing work in this area. We also provide an overview of the most important applications of Interactive Perception. We close this survey by discussing the remaining open questions. Thereby, we hope to define a field and inspire future work.

am

arXiv DOI Project Page [BibTex]

arXiv DOI Project Page [BibTex]


Thumb xl teaser
Optimizing Long-term Predictions for Model-based Policy Search

Doerr, A., Daniel, C., Nguyen-Tuong, D., Marco, A., Schaal, S., Toussaint, M., Trimpe, S.

Proceedings of 1st Annual Conference on Robot Learning (CoRL), 78, pages: 227-238, (Editors: Sergey Levine and Vincent Vanhoucke and Ken Goldberg), 1st Annual Conference on Robot Learning, November 2017 (conference)

Abstract
We propose a novel long-term optimization criterion to improve the robustness of model-based reinforcement learning in real-world scenarios. Learning a dynamics model to derive a solution promises much greater data-efficiency and reusability compared to model-free alternatives. In practice, however, modelbased RL suffers from various imperfections such as noisy input and output data, delays and unmeasured (latent) states. To achieve higher resilience against such effects, we propose to optimize a generative long-term prediction model directly with respect to the likelihood of observed trajectories as opposed to the common approach of optimizing a dynamics model for one-step-ahead predictions. We evaluate the proposed method on several artificial and real-world benchmark problems and compare it to PILCO, a model-based RL framework, in experiments on a manipulation robot. The results show that the proposed method is competitive compared to state-of-the-art model learning methods. In contrast to these more involved models, our model can directly be employed for policy search and outperforms a baseline method in the robot experiment.

am ics

PDF Project Page [BibTex]

PDF Project Page [BibTex]


Thumb xl qg net rev
Acquiring Target Stacking Skills by Goal-Parameterized Deep Reinforcement Learning

Li, W., Bohg, J., Fritz, M.

arXiv, November 2017 (article) Submitted

Abstract
Understanding physical phenomena is a key component of human intelligence and enables physical interaction with previously unseen environments. In this paper, we study how an artificial agent can autonomously acquire this intuition through interaction with the environment. We created a synthetic block stacking environment with physics simulation in which the agent can learn a policy end-to-end through trial and error. Thereby, we bypass to explicitly model physical knowledge within the policy. We are specifically interested in tasks that require the agent to reach a given goal state that may be different for every new trial. To this end, we propose a deep reinforcement learning framework that learns policies which are parametrized by a goal. We validated the model on a toy example navigating in a grid world with different target positions and in a block stacking task with different target structures of the final tower. In contrast to prior work, our policies show better generalization across different goals.

am

arXiv [BibTex]


no image
A New Data Source for Inverse Dynamics Learning

Kappler, D., Meier, F., Ratliff, N., Schaal, S.

In Proceedings IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, Piscataway, NJ, USA, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), September 2017 (inproceedings)

am

[BibTex]

[BibTex]


no image
Bayesian Regression for Artifact Correction in Electroencephalography

Fiebig, K., Jayaram, V., Hesse, T., Blank, A., Peters, J., Grosse-Wentrup, M.

Proceedings of the 7th Graz Brain-Computer Interface Conference 2017 - From Vision to Reality, pages: 131-136, (Editors: Müller-Putz G.R., Steyrl D., Wriessnegger S. C., Scherer R.), Graz University of Technology, Austria, Graz Brain-Computer Interface Conference, September 2017 (conference)

am ei

DOI [BibTex]

DOI [BibTex]


no image
Investigating Music Imagery as a Cognitive Paradigm for Low-Cost Brain-Computer Interfaces

Grossberger, L., Hohmann, M. R., Peters, J., Grosse-Wentrup, M.

Proceedings of the 7th Graz Brain-Computer Interface Conference 2017 - From Vision to Reality, pages: 160-164, (Editors: Müller-Putz G.R., Steyrl D., Wriessnegger S. C., Scherer R.), Graz University of Technology, Austria, Graz Brain-Computer Interface Conference, September 2017 (conference)

am ei

DOI [BibTex]

DOI [BibTex]


Thumb xl screen shot 2017 08 01 at 15.41.10
On the relevance of grasp metrics for predicting grasp success

Rubert, C., Kappler, D., Morales, A., Schaal, S., Bohg, J.

In Proceedings of the IEEE/RSJ International Conference of Intelligent Robots and Systems, September 2017 (inproceedings) Accepted

Abstract
We aim to reliably predict whether a grasp on a known object is successful before it is executed in the real world. There is an entire suite of grasp metrics that has already been developed which rely on precisely known contact points between object and hand. However, it remains unclear whether and how they may be combined into a general purpose grasp stability predictor. In this paper, we analyze these questions by leveraging a large scale database of simulated grasps on a wide variety of objects. For each grasp, we compute the value of seven metrics. Each grasp is annotated by human subjects with ground truth stability labels. Given this data set, we train several classification methods to find out whether there is some underlying, non-trivial structure in the data that is difficult to model manually but can be learned. Quantitative and qualitative results show the complexity of the prediction problem. We found that a good prediction performance critically depends on using a combination of metrics as input features. Furthermore, non-parametric and non-linear classifiers best capture the structure in the data.

am

Project Page [BibTex]

Project Page [BibTex]


no image
Local Bayesian Optimization of Motor Skills

Akrour, R., Sorokin, D., Peters, J., Neumann, G.

Proceedings of the 34th International Conference on Machine Learning, 70, pages: 41-50, Proceedings of Machine Learning Research, (Editors: Doina Precup, Yee Whye Teh), PMLR, International Conference on Machine Learning (ICML), August 2017 (conference)

am ei

link (url) Project Page [BibTex]

link (url) Project Page [BibTex]


Thumb xl pilqr cover
Combining Model-Based and Model-Free Updates for Trajectory-Centric Reinforcement Learning

Chebotar, Y., Hausman, K., Zhang, M., Sukhatme, G., Schaal, S., Levine, S.

Proceedings of the 34th International Conference on Machine Learning, 70, Proceedings of Machine Learning Research, (Editors: Doina Precup, Yee Whye Teh), PMLR, International Conference on Machine Learning (ICML), August 2017 (conference)

am

pdf video [BibTex]

pdf video [BibTex]


no image
Event-based State Estimation: An Emulation-based Approach

Trimpe, S.

IET Control Theory & Applications, 11(11):1684-1693, July 2017 (article)

Abstract
An event-based state estimation approach for reducing communication in a networked control system is proposed. Multiple distributed sensor agents observe a dynamic process and sporadically transmit their measurements to estimator agents over a shared bus network. Local event-triggering protocols ensure that data is transmitted only when necessary to meet a desired estimation accuracy. The event-based design is shown to emulate the performance of a centralised state observer design up to guaranteed bounds, but with reduced communication. The stability results for state estimation are extended to the distributed control system that results when the local estimates are used for feedback control. Results from numerical simulations and hardware experiments illustrate the effectiveness of the proposed approach in reducing network communication.

am ics

arXiv Supplementary material PDF DOI Project Page [BibTex]

arXiv Supplementary material PDF DOI Project Page [BibTex]


Thumb xl apollo system2 croped
Model-Based Policy Search for Automatic Tuning of Multivariate PID Controllers

Doerr, A., Nguyen-Tuong, D., Marco, A., Schaal, S., Trimpe, S.

In Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), pages: 5295-5301, IEEE, Piscataway, NJ, USA, IEEE International Conference on Robotics and Automation (ICRA), May 2017 (inproceedings)

am ics

PDF arXiv DOI Project Page [BibTex]

PDF arXiv DOI Project Page [BibTex]


Thumb xl learning ct block diagram v2
Learning Feedback Terms for Reactive Planning and Control

Rai, A., Sutanto, G., Schaal, S., Meier, F.

Proceedings 2017 IEEE International Conference on Robotics and Automation (ICRA), IEEE, Piscataway, NJ, USA, IEEE International Conference on Robotics and Automation (ICRA), May 2017 (conference)

am

pdf video [BibTex]

pdf video [BibTex]


Thumb xl this one
Virtual vs. Real: Trading Off Simulations and Physical Experiments in Reinforcement Learning with Bayesian Optimization

Marco, A., Berkenkamp, F., Hennig, P., Schoellig, A. P., Krause, A., Schaal, S., Trimpe, S.

In Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), pages: 1557-1563, IEEE, Piscataway, NJ, USA, IEEE International Conference on Robotics and Automation (ICRA), May 2017 (inproceedings)

am ics pn

PDF arXiv ICRA 2017 Spotlight presentation Virtual vs. Real - Video explanation DOI Project Page [BibTex]

PDF arXiv ICRA 2017 Spotlight presentation Virtual vs. Real - Video explanation DOI Project Page [BibTex]


Thumb xl fig  quali  arm
Probabilistic Articulated Real-Time Tracking for Robot Manipulation

(Best Paper of RA-L 2017, Finalist of Best Robotic Vision Paper Award of ICRA 2017)

Garcia Cifuentes, C., Issac, J., Wüthrich, M., Schaal, S., Bohg, J.

IEEE Robotics and Automation Letters (RA-L), 2(2):577-584, April 2017 (article)

Abstract
We propose a probabilistic filtering method which fuses joint measurements with depth images to yield a precise, real-time estimate of the end-effector pose in the camera frame. This avoids the need for frame transformations when using it in combination with visual object tracking methods. Precision is achieved by modeling and correcting biases in the joint measurements as well as inaccuracies in the robot model, such as poor extrinsic camera calibration. We make our method computationally efficient through a principled combination of Kalman filtering of the joint measurements and asynchronous depth-image updates based on the Coordinate Particle Filter. We quantitatively evaluate our approach on a dataset recorded from a real robotic platform, annotated with ground truth from a motion capture system. We show that our approach is robust and accurate even under challenging conditions such as fast motion, significant and long-term occlusions, and time-varying biases. We release the dataset along with open-source code of our approach to allow for quantitative comparison with alternative approaches.

am

arXiv video code and dataset video PDF DOI Project Page [BibTex]


no image
Anticipatory Action Selection for Human-Robot Table Tennis

Wang, Z., Boularias, A., Mülling, K., Schölkopf, B., Peters, J.

Artificial Intelligence, 247, pages: 399-414, 2017, Special Issue on AI and Robotics (article)

Abstract
Abstract Anticipation can enhance the capability of a robot in its interaction with humans, where the robot predicts the humans' intention for selecting its own action. We present a novel framework of anticipatory action selection for human-robot interaction, which is capable to handle nonlinear and stochastic human behaviors such as table tennis strokes and allows the robot to choose the optimal action based on prediction of the human partner's intention with uncertainty. The presented framework is generic and can be used in many human-robot interaction scenarios, for example, in navigation and human-robot co-manipulation. In this article, we conduct a case study on human-robot table tennis. Due to the limited amount of time for executing hitting movements, a robot usually needs to initiate its hitting movement before the opponent hits the ball, which requires the robot to be anticipatory based on visual observation of the opponent's movement. Previous work on Intention-Driven Dynamics Models (IDDM) allowed the robot to predict the intended target of the opponent. In this article, we address the problem of action selection and optimal timing for initiating a chosen action by formulating the anticipatory action selection as a Partially Observable Markov Decision Process (POMDP), where the transition and observation are modeled by the \{IDDM\} framework. We present two approaches to anticipatory action selection based on the \{POMDP\} formulation, i.e., a model-free policy learning method based on Least-Squares Policy Iteration (LSPI) that employs the \{IDDM\} for belief updates, and a model-based Monte-Carlo Planning (MCP) method, which benefits from the transition and observation model by the IDDM. Experimental results using real data in a simulated environment show the importance of anticipatory action selection, and that \{POMDPs\} are suitable to formulate the anticipatory action selection problem by taking into account the uncertainties in prediction. We also show that existing algorithms for POMDPs, such as \{LSPI\} and MCP, can be applied to substantially improve the robot's performance in its interaction with humans.

am ei

DOI Project Page [BibTex]

DOI Project Page [BibTex]


no image
Robot Learning

Peters, J., Lee, D., Kober, J., Nguyen-Tuong, D., Bagnell, J., Schaal, S.

In Springer Handbook of Robotics, pages: 357-394, 15, 2nd, (Editors: Siciliano, Bruno and Khatib, Oussama), Springer International Publishing, 2017 (inbook)

am ei

Project Page [BibTex]

Project Page [BibTex]

2006


no image
Learning operational space control

Peters, J., Schaal, S.

In Robotics: Science and Systems II (RSS 2006), pages: 255-262, (Editors: Gaurav S. Sukhatme and Stefan Schaal and Wolfram Burgard and Dieter Fox), Cambridge, MA: MIT Press, RSS , 2006, clmc (inproceedings)

Abstract
While operational space control is of essential importance for robotics and well-understood from an analytical point of view, it can be prohibitively hard to achieve accurate control in face of modeling errors, which are inevitable in complex robots, e.g., humanoid robots. In such cases, learning control methods can offer an interesting alternative to analytical control algorithms. However, the resulting learning problem is ill-defined as it requires to learn an inverse mapping of a usually redundant system, which is well known to suffer from the property of non-covexity of the solution space, i.e., the learning system could generate motor commands that try to steer the robot into physically impossible configurations. A first important insight for this paper is that, nevertheless, a physically correct solution to the inverse problem does exits when learning of the inverse map is performed in a suitable piecewise linear way. The second crucial component for our work is based on a recent insight that many operational space controllers can be understood in terms of a constraint optimal control problem. The cost function associated with this optimal control problem allows us to formulate a learning algorithm that automatically synthesizes a globally consistent desired resolution of redundancy while learning the operational space controller. From the view of machine learning, the learning problem corresponds to a reinforcement learning problem that maximizes an immediate reward and that employs an expectation-maximization policy search algorithm. Evaluations on a three degrees of freedom robot arm illustrate the feasability of our suggested approach.

am ei

link (url) [BibTex]

2006


link (url) [BibTex]


no image
Reinforcement Learning for Parameterized Motor Primitives

Peters, J., Schaal, S.

In Proceedings of the 2006 International Joint Conference on Neural Networks, pages: 73-80, IJCNN, 2006, clmc (inproceedings)

Abstract
One of the major challenges in both action generation for robotics and in the understanding of human motor control is to learn the "building blocks of movement generation", called motor primitives. Motor primitives, as used in this paper, are parameterized control policies such as splines or nonlinear differential equations with desired attractor properties. While a lot of progress has been made in teaching parameterized motor primitives using supervised or imitation learning, the self-improvement by interaction of the system with the environment remains a challenging problem. In this paper, we evaluate different reinforcement learning approaches for improving the performance of parameterized motor primitives. For pursuing this goal, we highlight the difficulties with current reinforcement learning methods, and outline both established and novel algorithms for the gradient-based improvement of parameterized policies. We compare these algorithms in the context of motor primitive learning, and show that our most modern algorithm, the Episodic Natural Actor-Critic outperforms previous algorithms by at least an order of magnitude. We demonstrate the efficiency of this reinforcement learning method in the application of learning to hit a baseball with an anthropomorphic robot arm.

am ei

link (url) DOI [BibTex]

link (url) DOI [BibTex]


no image
Statistical Learning of LQG controllers

Theodorou, E.

Technical Report-2006-1, Computational Action and Vision Lab University of Minnesota, 2006, clmc (techreport)

am

PDF [BibTex]

PDF [BibTex]


no image
Approximate nearest neighbor regression in very high dimensions

Vijayakumar, S., DSouza, A., Schaal, S.

In Nearest-Neighbor Methods in Learning and Vision, pages: 103-142, (Editors: Shakhnarovich, G.;Darrell, T.;Indyk, P.), Cambridge, MA: MIT Press, 2006, clmc (inbook)

am

link (url) [BibTex]

link (url) [BibTex]

2003


no image
Dynamic movement primitives - A framework for motor control in humans and humanoid robots

Schaal, S.

In The International Symposium on Adaptive Motion of Animals and Machines, Kyoto, Japan, March 4-8, 2003, March 2003, clmc (inproceedings)

Abstract
Sensory-motor integration is one of the key issues in robotics. In this paper, we propose an approach to rhythmic arm movement control that is synchronized with an external signal based on exploiting a simple neural oscillator network. Trajectory generation by the neural oscillator is a biologically inspired method that can allow us to generate a smooth and continuous trajectory. The parameter tuning of the oscillators is used to generate a synchronized movement with wide intervals. We adopted the method for the drumming task as an example task. By using this method, the robot can realize synchronized drumming with wide drumming intervals in real time. The paper also shows the experimental results of drumming by a humanoid robot.

am

link (url) [BibTex]

2003


link (url) [BibTex]


no image
Bayesian backfitting

D’Souza, A., Vijayakumar, S., Schaal, S.

In Proceedings of the 10th Joint Symposium on Neural Computation (JSNC 2003), Irvine, CA, May 2003, 2003, clmc (inproceedings)

Abstract
We present an algorithm aimed at addressing both computational and analytical intractability of Bayesian regression models which operate in very high-dimensional, usually underconstrained spaces. Several domains of research frequently provide such datasets, including chemometrics [2], and human movement analysis [1]. The literature in nonparametric statistics provides interesting solutions such as Backfitting [3] and Partial Least Squares [4], which are extremely robust and efficient, yet lack a probabilistic interpretation that could place them in the context of current research in statistical learning algorithms that emphasize the estimation of confidence, posterior distributions, and model complexity. In order to achieve numerical robustness and low computational cost, we first derive a novel Bayesian interpretation of Backfitting (BB) as a computationally efficient regression algorithm. BBÕs learning complexity scales linearly with the input dimensionality by decoupling inference among individual input dimensions. We embed BB in an efficient, locally variational model selection mechanism that automatically grows the number of backfitting experts in a mixture-of-experts regression model. We demonstrate the effectiveness of the algorithm in performing principled regularization of model complexity when fitting nonlinear manifolds while avoiding the numerical hazards associated with highly underconstrained problems. We also note that this algorithm appears applicable in various areas of neural computation, e.g., in abstract models of computational neuroscience, or implementations of statistical learning on artificial systems.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Reinforcement learning for humanoid robotics

Peters, J., Vijayakumar, S., Schaal, S.

In IEEE-RAS International Conference on Humanoid Robots (Humanoids2003), Karlsruhe, Germany, Sept.29-30, 2003, clmc (inproceedings)

Abstract
Reinforcement learning offers one of the most general framework to take traditional robotics towards true autonomy and versatility. However, applying reinforcement learning to high dimensional movement systems like humanoid robots remains an unsolved problem. In this paper, we discuss different approaches of reinforcement learning in terms of their applicability in humanoid robotics. Methods can be coarsely classified into three different categories, i.e., greedy methods, `vanilla' policy gradient methods, and natural gradient methods. We discuss that greedy methods are not likely to scale into the domain humanoid robotics as they are problematic when used with function approximation. `Vanilla' policy gradient methods on the other hand have been successfully applied on real-world robots including at least one humanoid robot. We demonstrate that these methods can be significantly improved using the natural policy gradient instead of the regular policy gradient. A derivation of the natural policy gradient is provided, proving that the average policy gradient of Kakade (2002) is indeed the true natural gradient. A general algorithm for estimating the natural gradient, the Natural Actor-Critic algorithm, is introduced. This algorithm converges to the nearest local minimum of the cost function with respect to the Fisher information metric under suitable conditions. The algorithm outperforms non-natural policy gradients by far in a cart-pole balancing evaluation, and for learning nonlinear dynamic motor primitives for humanoid robot control. It offers a promising route for the development of reinforcement learning for truly high dimensionally continuous state-action systems.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Discovering imitation strategies through categorization of multi-cimensional data

Billard, A., Epars, Y., Schaal, S., Cheng, G.

In IEEE International Conference on Intelligent Robots and Systems (IROS 2003), Las Vegas, NV, Oct. 27-31, 2003, clmc (inproceedings)

Abstract
An essential problem of imitation is that of determining Ówhat to imitateÓ, i.e. to determine which of the many features of the demonstration are relevant to the task and which should be reproduced. The strategy followed by the imitator can be modeled as a hierarchical optimization system, which minimizes the discrepancy between two multidimensional datasets. We consider imitation of a manipulation task. To classify across manipulation strategies, we apply a probabilistic analysis to data in Cartesian and joint spaces. We determine a general metric that optimizes the policy of task reproduction, following strategy determination. The model successfully discovers strategies in six different manipulation tasks and controls task reproduction by a full body humanoid robot. or the complete path followed by the demonstrator. We follow a similar taxonomy and apply it to the learning and reproduction of a manipulation task by a humanoid robot. We take the perspective that the features of the movements to imitate are those that appear most frequently, i.e. the invariants in time. The model builds upon previous work [3], [4] and is composed of a hierarchical time delay neural network that extracts invariant features from a manipulation task performed by a human demonstrator. The system analyzes the Carthesian trajectories of the objects and the joint

am

link (url) [BibTex]

link (url) [BibTex]


no image
Scaling reinforcement learning paradigms for motor learning

Peters, J., Vijayakumar, S., Schaal, S.

In Proceedings of the 10th Joint Symposium on Neural Computation (JSNC 2003), Irvine, CA, May 2003, 2003, clmc (inproceedings)

Abstract
Reinforcement learning offers a general framework to explain reward related learning in artificial and biological motor control. However, current reinforcement learning methods rarely scale to high dimensional movement systems and mainly operate in discrete, low dimensional domains like game-playing, artificial toy problems, etc. This drawback makes them unsuitable for application to human or bio-mimetic motor control. In this poster, we look at promising approaches that can potentially scale and suggest a novel formulation of the actor-critic algorithm which takes steps towards alleviating the current shortcomings. We argue that methods based on greedy policies are not likely to scale into high-dimensional domains as they are problematic when used with function approximation Ð a must when dealing with continuous domains. We adopt the path of direct policy gradient based policy improvements since they avoid the problems of unstabilizing dynamics encountered in traditional value iteration based updates. While regular policy gradient methods have demonstrated promising results in the domain of humanoid notor control, we demonstrate that these methods can be significantly improved using the natural policy gradient instead of the regular policy gradient. Based on this, it is proved that KakadeÕs Ôaverage natural policy gradientÕ is indeed the true natural gradient. A general algorithm for estimating the natural gradient, the Natural Actor-Critic algorithm, is introduced. This algorithm converges with probability one to the nearest local minimum in Riemannian space of the cost function. The algorithm outperforms nonnatural policy gradients by far in a cart-pole balancing evaluation, and offers a promising route for the development of reinforcement learning for truly high-dimensionally continuous state-action systems.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Learning attractor landscapes for learning motor primitives

Ijspeert, A., Nakanishi, J., Schaal, S.

In Advances in Neural Information Processing Systems 15, pages: 1547-1554, (Editors: Becker, S.;Thrun, S.;Obermayer, K.), Cambridge, MA: MIT Press, 2003, clmc (inproceedings)

Abstract
If globally high dimensional data has locally only low dimensional distributions, it is advantageous to perform a local dimensionality reduction before further processing the data. In this paper we examine several techniques for local dimensionality reduction in the context of locally weighted linear regression. As possible candidates, we derive local versions of factor analysis regression, principle component regression, principle component regression on joint distributions, and partial least squares regression. After outlining the statistical bases of these methods, we perform Monte Carlo simulations to evaluate their robustness with respect to violations of their statistical assumptions. One surprising outcome is that locally weighted partial least squares regression offers the best average results, thus outperforming even factor analysis, the theoretically most appealing of our candidate techniques.Ê

am

link (url) [BibTex]

link (url) [BibTex]


no image
Learning from demonstration and adaptation of biped locomotion with dynamical movement primitives

Nakanishi, J., Morimoto, J., Endo, G., Schaal, S., Kawato, M.

In Workshop on Robot Learning by Demonstration, IEEE International Conference on Intelligent Robots and Systems (IROS 2003), Las Vegas, NV, Oct. 27-31, 2003, clmc (inproceedings)

Abstract
In this paper, we report on our research for learning biped locomotion from human demonstration. Our ultimate goal is to establish a design principle of a controller in order to achieve natural human-like locomotion. We suggest dynamical movement primitives as a CPG of a biped robot, an approach we have previously proposed for learning and encoding complex human movements. Demonstrated trajectories are learned through the movement primitives by locally weighted regression, and the frequency of the learned trajectories is adjusted automatically by a novel frequency adaptation algorithm based on phase resetting and entrainment of oscillators. Numerical simulations demonstrate the effectiveness of the proposed locomotion controller.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Movement planning and imitation by shaping nonlinear attractors

Schaal, S.

In Proceedings of the 12th Yale Workshop on Adaptive and Learning Systems, Yale University, New Haven, CT, 2003, clmc (inproceedings)

Abstract
Given the continuous stream of movements that biological systems exhibit in their daily activities, an account for such versatility and creativity has to assume that movement sequences consist of segments, executed either in sequence or with partial or complete overlap. Therefore, a fundamental question that has pervaded research in motor control both in artificial and biological systems revolves around identifying movement primitives (a.k.a. units of actions, basis behaviors, motor schemas, etc.). What are the fundamental building blocks that are strung together, adapted to, and created for ever new behaviors? This paper summarizes results that led to the hypothesis of Dynamic Movement Primitives (DMP). DMPs are units of action that are formalized as stable nonlinear attractor systems. They are useful for autonomous robotics as they are highly flexible in creating complex rhythmic (e.g., locomotion) and discrete (e.g., a tennis swing) behaviors that can quickly be adapted to the inevitable perturbations of a dy-namically changing, stochastic environment. Moreover, DMPs provide a formal framework that also lends itself to investigations in computational neuroscience. A recent finding that allows creating DMPs with the help of well-understood statistical learning methods has elevated DMPs from a more heuristic to a principled modeling approach, and, moreover, created a new foundation for imitation learning. Theoretical insights, evaluations on a humanoid robot, and behavioral and brain imaging data will serve to outline the framework of DMPs for a general approach to motor control and imitation in robotics and biology.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Computational approaches to motor learning by imitation

Schaal, S., Ijspeert, A., Billard, A.

Philosophical Transaction of the Royal Society of London: Series B, Biological Sciences, 358(1431):537-547, 2003, clmc (article)

Abstract
Movement imitation requires a complex set of mechanisms that map an observed movement of a teacher onto one's own movement apparatus. Relevant problems include movement recognition, pose estimation, pose tracking, body correspondence, coordinate transformation from external to egocentric space, matching of observed against previously learned movement, resolution of redundant degrees-of-freedom that are unconstrained by the observation, suitable movement representations for imitation, modularization of motor control, etc. All of these topics by themselves are active research problems in computational and neurobiological sciences, such that their combination into a complete imitation system remains a daunting undertaking - indeed, one could argue that we need to understand the complete perception-action loop. As a strategy to untangle the complexity of imitation, this paper will examine imitation purely from a computational point of view, i.e. we will review statistical and mathematical approaches that have been suggested for tackling parts of the imitation problem, and discuss their merits, disadvantages and underlying principles. Given the focus on action recognition of other contributions in this special issue, this paper will primarily emphasize the motor side of imitation, assuming that a perceptual system has already identified important features of a demonstrated movement and created their corresponding spatial information. Based on the formalization of motor control in terms of control policies and their associated performance criteria, useful taxonomies of imitation learning can be generated that clarify different approaches and future research directions.

am

link (url) [BibTex]

link (url) [BibTex]

2000


no image
Reciprocal excitation between biological and robotic research

Schaal, S., Sternad, D., Dean, W., Kotoska, S., Osu, R., Kawato, M.

In Sensor Fusion and Decentralized Control in Robotic Systems III, Proceedings of SPIE, 4196, pages: 30-40, Boston, MA, Nov.5-8, 2000, November 2000, clmc (inproceedings)

Abstract
While biological principles have inspired researchers in computational and engineering research for a long time, there is still rather limited knowledge flow back from computational to biological domains. This paper presents examples of our work where research on anthropomorphic robots lead us to new insights into explaining biological movement phenomena, starting from behavioral studies up to brain imaging studies. Our research over the past years has focused on principles of trajectory formation with nonlinear dynamical systems, on learning internal models for nonlinear control, and on advanced topics like imitation learning. The formal and empirical analyses of the kinematics and dynamics of movements systems and the tasks that they need to perform lead us to suggest principles of motor control that later on we found surprisingly related to human behavior and even brain activity.

am

link (url) [BibTex]

2000


link (url) [BibTex]


no image
Nonlinear dynamical systems as movement primitives

Schaal, S., Kotosaka, S., Sternad, D.

In Humanoids2000, First IEEE-RAS International Conference on Humanoid Robots, CD-Proceedings, Cambridge, MA, September 2000, clmc (inproceedings)

Abstract
This paper explores the idea to create complex human-like movements from movement primitives based on nonlinear attractor dynamics. Each degree-of-freedom of a limb is assumed to have two independent abilities to create movement, one through a discrete dynamic system, and one through a rhythmic system. The discrete system creates point-to-point movements based on internal or external target specifications. The rhythmic system can add an additional oscillatory movement relative to the current position of the discrete system. In the present study, we develop appropriate dynamic systems that can realize the above model, motivate the particular choice of the systems from a biological and engineering point of view, and present simulation results of the performance of such movement primitives. The model was implemented for a drumming task on a humanoid robot

am

link (url) [BibTex]

link (url) [BibTex]


no image
Real Time Learning in Humanoids: A challenge for scalability of Online Algorithms

Vijayakumar, S., Schaal, S.

In Humanoids2000, First IEEE-RAS International Conference on Humanoid Robots, CD-Proceedings, Cambridge, MA, September 2000, clmc (inproceedings)

Abstract
While recent research in neural networks and statistical learning has focused mostly on learning from finite data sets without stringent constraints on computational efficiency, there is an increasing number of learning problems that require real-time performance from an essentially infinite stream of incrementally arriving data. This paper demonstrates how even high-dimensional learning problems of this kind can successfully be dealt with by techniques from nonparametric regression and locally weighted learning. As an example, we describe the application of one of the most advanced of such algorithms, Locally Weighted Projection Regression (LWPR), to the on-line learning of the inverse dynamics model of an actual seven degree-of-freedom anthropomorphic robot arm. LWPR's linear computational complexity in the number of input dimensions, its inherent mechanisms of local dimensionality reduction, and its sound learning rule based on incremental stochastic leave-one-out cross validation allows -- to our knowledge for the first time -- implementing inverse dynamics learning for such a complex robot with real-time performance. In our sample task, the robot acquires the local inverse dynamics model needed to trace a figure-8 in only 60 seconds of training.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Synchronized robot drumming by neural oscillator

Kotosaka, S., Schaal, S.

In The International Symposium on Adaptive Motion of Animals and Machines, Montreal, Canada, August 2000, clmc (inproceedings)

Abstract
Sensory-motor integration is one of the key issues in robotics. In this paper, we propose an approach to rhythmic arm movement control that is synchronized with an external signal based on exploiting a simple neural oscillator network. Trajectory generation by the neural oscillator is a biologically inspired method that can allow us to generate a smooth and continuous trajectory. The parameter tuning of the oscillators is used to generate a synchronized movement with wide intervals. We adopted the method for the drumming task as an example task. By using this method, the robot can realize synchronized drumming with wide drumming intervals in real time. The paper also shows the experimental results of drumming by a humanoid robot.

am

link (url) [BibTex]

link (url) [BibTex]


no image
A brachiating robot controller

Nakanishi, J., Fukuda, T., Koditschek, D. E.

IEEE Transactions on Robotics and Automation, 16(2):109-123, 2000, clmc (article)

Abstract
We report on our empirical studies of a new controller for a two-link brachiating robot. Motivated by the pendulum-like motion of an apeâ??s brachiation, we encode this task as the output of a â??target dynamical system.â? Numerical simulations indicate that the resulting controller solves a number of brachiation problems that we term the â??ladder,â? â??swing-up,â? and â??ropeâ? problems. Preliminary analysis provides some explanation for this success. The proposed controller is implemented on a physical system in our laboratory. The robot achieves behaviors including â??swing locomotionâ? and â??swing upâ? and is capable of continuous locomotion over several rungs of a ladder. We discuss a number of formal questions whose answers will be required to gain a full understanding of the strengths and weaknesses of this approach.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Real-time robot learning with locally weighted statistical learning

Schaal, S., Atkeson, C. G., Vijayakumar, S.

In International Conference on Robotics and Automation (ICRA2000), San Francisco, April 2000, 2000, clmc (inproceedings)

Abstract
Locally weighted learning (LWL) is a class of statistical learning techniques that provides useful representations and training algorithms for learning about complex phenomena during autonomous adaptive control of robotic systems. This paper introduces several LWL algorithms that have been tested successfully in real-time learning of complex robot tasks. We discuss two major classes of LWL, memory-based LWL and purely incremental LWL that does not need to remember any data explicitly. In contrast to the traditional beliefs that LWL methods cannot work well in high-dimensional spaces, we provide new algorithms that have been tested in up to 50 dimensional learning problems. The applicability of our LWL algorithms is demonstrated in various robot learning examples, including the learning of devil-sticking, pole-balancing of a humanoid robot arm, and inverse-dynamics learning for a seven degree-of-freedom robot.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Biomimetic gaze stabilization

Shibata, T., Schaal, S.

In Robot learning: an Interdisciplinary approach, pages: 31-52, (Editors: Demiris, J.;Birk, A.), World Scientific, 2000, clmc (inbook)

Abstract
Accurate oculomotor control is one of the essential pre-requisites for successful visuomotor coordination. In this paper, we suggest a biologically inspired control system for learning gaze stabilization with a biomimetic robotic oculomotor system. In a stepwise fashion, we develop a control circuit for the vestibulo-ocular reflex (VOR) and the opto-kinetic response (OKR), and add a nonlinear learning network to allow adaptivity. We discuss the parallels and differences of our system with biological oculomotor control and suggest solutions how to deal with nonlinearities and time delays in the control system. In simulation and actual robot studies, we demonstrate that our system can learn gaze stabilization in real time in only a few seconds with high final accuracy.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Fast learning of biomimetic oculomotor control with nonparametric regression networks

Shibata, T., Schaal, S.

In International Conference on Robotics and Automation (ICRA2000), pages: 3847-3854, San Francisco, April 2000, 2000, clmc (inproceedings)

Abstract
Accurate oculomotor control is one of the essential pre-requisites of successful visuomotor coordination. Given the variable nonlinearities of the geometry of binocular vision as well as the possible nonlinearities of the oculomotor plant, it is desirable to accomplish accurate oculomotor control through learning approaches. In this paper, we investigate learning control for a biomimetic active vision system mounted on a humanoid robot. By combining a biologically inspired cerebellar learning scheme with a state-of-the-art statistical learning network, our robot system is able to acquire high performance visual stabilization reflexes after about 40 seconds of learning despite significant nonlinearities and processing delays in the system.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Interaction of rhythmic and discrete pattern generators in single joint movements

Sternad, D., Dean, W. J., Schaal, S.

Human Movement Science, 19(4):627-665, 2000, clmc (article)

Abstract
The study investigates a single-joint movement task that combines a translatory and cyclic component with the objective to investigate the interaction of discrete and rhythmic movement elements. Participants performed an elbow movement in the horizontal plane, oscillating at a prescribed frequency around one target and shifting to a second target upon a trigger signal, without stopping the oscillation. Analyses focused on extracting the mutual influences of the rhythmic and the discrete component of the task. Major findings are: (1) The onset of the discrete movement was confined to a limited phase window in the rhythmic cycle. (2) Its duration was influenced by the period of oscillation. (3) The rhythmic oscillation was "perturbed" by the discrete movement as indicated by phase resetting. On the basis of these results we propose a model for the coordination of discrete and rhythmic actions (K. Matsuoka, Sustained oscillations generated by mutually inhibiting neurons with adaptations, Biological Cybernetics 52 (1985) 367-376; Mechanisms of frequency and pattern control in the neural rhythm generators, Biological Cybernetics 56 (1987) 345-353). For rhythmic movements an oscillatory pattern generator is developed following models of half-center oscillations (D. Bullock, S. Grossberg, The VITE model: a neural command circuit for generating arm and articulated trajectories, in: J.A.S. Kelso, A.J. Mandel, M. F. Shlesinger (Eds.), Dynamic Patterns in Complex Systems. World Scientific. Singapore. 1988. pp. 305-326). For discrete movements a point attractor dynamics is developed close to the VITE model For each joint degree of freedom both pattern generators co-exist but exert mutual inhibition onto each other. The suggested modeling framework provides a unified account for both discrete and rhythmic movements on the basis of neuronal circuitry. Simulation results demonstrated that the effects observed in human performance can be replicated using the two pattern generators with a mutually inhibiting coupling.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Locally weighted projection regression: An O(n) algorithm for incremental real time learning in high dimensional spaces

Vijayakumar, S., Schaal, S.

In Proceedings of the Seventeenth International Conference on Machine Learning (ICML 2000), 1, pages: 288-293, Stanford, CA, 2000, clmc (inproceedings)

Abstract
Locally weighted projection regression is a new algorithm that achieves nonlinear function approximation in high dimensional spaces with redundant and irrelevant input dimensions. At its core, it uses locally linear models, spanned by a small number of univariate regressions in selected directions in input space. This paper evaluates different methods of projection regression and derives a nonlinear function approximator based on them. This nonparametric local learning system i) learns rapidly with second order learning methods based on incremental training, ii) uses statistically sound stochastic cross validation to learn iii) adjusts its weighting kernels based on local information only, iv) has a computational complexity that is linear in the number of inputs, and v) can deal with a large number of - possibly redundant - inputs, as shown in evaluations with up to 50 dimensional data sets. To our knowledge, this is the first truly incremental spatially localized learning method to combine all these properties.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Dynamics of a bouncing ball in human performance

Sternad, D., Duarte, M., Katsumata, H., Schaal, S.

Physical Review E, 63(011902):1-8, 2000, clmc (article)

Abstract
On the basis of a modified bouncing-ball model, we investigated whether human movements utilize principles of dynamic stability in their performance of a similar movement task. Stability analyses of the model provided predictions about conditions indicative of a dynamically stable period-one regime. In a series of experiments, human subjects bounced a ball rhythmically on a racket and displayed these conditions supporting that they attuned to and exploited the dynamic stability properties of the task.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Inverse kinematics for humanoid robots

Tevatia, G., Schaal, S.

In International Conference on Robotics and Automation (ICRA2000), pages: 294-299, San Fransisco, April 24-28, 2000, 2000, clmc (inproceedings)

Abstract
Real-time control of the endeffector of a humanoid robot in external coordinates requires computationally efficient solutions of the inverse kinematics problem. In this context, this paper investigates methods of resolved motion rate control (RMRC) that employ optimization criteria to resolve kinematic redundancies. In particular we focus on two established techniques, the pseudo inverse with explicit optimization and the extended Jacobian method. We prove that the extended Jacobian method includes pseudo-inverse methods as a special solution. In terms of computational complexity, however, pseudo-inverse and extended Jacobian differ significantly in favor of pseudo-inverse methods. Employing numerical estimation techniques, we introduce a computationally efficient version of the extended Jacobian with performance comparable to the original version . Our results are illustrated in simulation studies with a multiple degree-of-freedom robot, and were tested on a 30 degree-of-freedom robot. 

am

link (url) [BibTex]

link (url) [BibTex]


no image
Fast and efficient incremental learning for high-dimensional movement systems

Vijayakumar, S., Schaal, S.

In International Conference on Robotics and Automation (ICRA2000), San Francisco, April 2000, 2000, clmc (inproceedings)

Abstract
We introduce a new algorithm, Locally Weighted Projection Regression (LWPR), for incremental real-time learning of nonlinear functions, as particularly useful for problems of autonomous real-time robot control that re-quires internal models of dynamics, kinematics, or other functions. At its core, LWPR uses locally linear models, spanned by a small number of univariate regressions in selected directions in input space, to achieve piecewise linear function approximation. The most outstanding properties of LWPR are that it i) learns rapidly with second order learning methods based on incremental training, ii) uses statistically sound stochastic cross validation to learn iii) adjusts its local weighting kernels based on only local information to avoid interference problems, iv) has a computational complexity that is linear in the number of inputs, and v) can deal with a large number ofâ??possibly redundant and/or irrelevantâ??inputs, as shown in evaluations with up to 50 dimensional data sets for learning the inverse dynamics of an anthropomorphic robot arm. To our knowledge, this is the first incremental neural network learning method to combine all these properties and that is well suited for complex on-line learning problems in robotics.

am

link (url) [BibTex]

link (url) [BibTex]


no image
On-line learning for humanoid robot systems

Conradt, J., Tevatia, G., Vijayakumar, S., Schaal, S.

In Proceedings of the Seventeenth International Conference on Machine Learning (ICML 2000), 1, pages: 191-198, Stanford, CA, 2000, clmc (inproceedings)

Abstract
Humanoid robots are high-dimensional movement systems for which analytical system identification and control methods are insufficient due to unknown nonlinearities in the system structure. As a way out, supervised learning methods can be employed to create model-based nonlinear controllers which use functions in the control loop that are estimated by learning algorithms. However, internal models for humanoid systems are rather high-dimensional such that conventional learning algorithms would suffer from slow learning speed, catastrophic interference, and the curse of dimensionality. In this paper we explore a new statistical learning algorithm, locally weighted projection regression (LWPR), for learning internal models in real-time. LWPR is a nonparametric spatially localized learning system that employs the less familiar technique of partial least squares regression to represent functional relationships in a piecewise linear fashion. The algorithm can work successfully in very high dimensional spaces and detect irrelevant and redundant inputs while only requiring a computational complexity that is linear in the number of input dimensions. We demonstrate the application of the algorithm in learning two classical internal models of robot control, the inverse kinematics and the inverse dynamics of an actual seven degree-of-freedom anthropomorphic robot arm. For both examples, LWPR can achieve excellent real-time learning results from less than one hour of actual training data.

am

link (url) [BibTex]

link (url) [BibTex]


no image
Humanoid Robot DB

Kotosaka, S., Shibata, T., Schaal, S.

In Proceedings of the International Conference on Machine Automation (ICMA2000), pages: 21-26, 2000, clmc (inproceedings)

am

[BibTex]

[BibTex]

1993


no image
Learning passive motor control strategies with genetic algorithms

Schaal, S., Sternad, D.

In 1992 Lectures in complex systems, pages: 913-918, (Editors: Nadel, L.;Stein, D.), Addison-Wesley, Redwood City, CA, 1993, clmc (inbook)

Abstract
This study investigates learning passive motor control strategies. Passive control is understood as control without active error correction; the movement is stabilized by particular properties of the controlling dynamics. We analyze the task of juggling a ball on a racket. An approximation to the optimal solution of the task is derived by means of optimization theory. In order to model the learning process, the problem is coded for a genetic algorithm in representations without sensory or with sensory information. For all representations the genetic algorithm is able to find passive control strategies, but learning speed and the quality of the outcome are significantly different. A comparison with data from human subjects shows that humans seem to apply yet different movement strategies to the ones proposed. For the feedback representation some implications arise for learning from demonstration.

am

link (url) [BibTex]

1993


link (url) [BibTex]


no image
A genetic algorithm for evolution from an ecological perspective

Sternad, D., Schaal, S.

In 1992 Lectures in Complex Systems, pages: 223-231, (Editors: Nadel, L.;Stein, D.), Addison-Wesley, Redwood City, CA, 1993, clmc (inbook)

Abstract
In the population model presented, an evolutionary dynamic is explored which is based on the operator characteristics of genetic algorithms. An essential modification in the genetic algorithms is the inclusion of a constraint in the mixing of the gene pool. The pairing for the crossover is governed by a selection principle based on a complementarity criterion derived from the theoretical tenet of perception-action (P-A) mutuality of ecological psychology. According to Swenson and Turvey [37] P-A mutuality underlies evolution and is an integral part of its thermodynamics. The present simulation tested the contribution of P-A-cycles in evolutionary dynamics. A numerical experiment compares the population's evolution with and without this intentional component. The effect is measured in the difference of the rate of energy dissipation, as well as in three operationalized aspects of complexity. The results support the predicted increase in the rate of energy dissipation, paralleled by an increase in the average heterogeneity of the population. Furthermore, the spatio-temporal evolution of the system is tested for the characteristic power-law relations of a nonlinear system poised in a critical state. The frequency distribution of consecutive increases in population size shows a significantly different exponent in functional relationship.

am

[BibTex]

[BibTex]


no image
Roles for memory-based learning in robotics

Atkeson, C. G., Schaal, S.

In Proceedings of the Sixth International Symposium on Robotics Research, pages: 503-521, Hidden Valley, PA, 1993, clmc (inproceedings)

am

[BibTex]

[BibTex]


no image
Design concurrent calculation: A CAD- and data-integrated approach

Schaal, S., Ehrlenspiel, K.

Journal of Engineering Design, 4, pages: 71-85, 1993, clmc (article)

Abstract
Besides functional regards, product design demands increasingly more for further reaching considerations. Quality alone cannot suffice anymore to compete in the market; design for manufacturability, for assembly, for recycling, etc., are well-known keywords. Those can largely be reduced to the necessity of design for costs. This paper focuses on a CAD-based approach to design concurrent calculation. It will discuss how, in the meantime well-established, tools like feature technology, knowledge-based systems, and relational databases can be blended into one coherent concept to achieve an entirely CAD- and data-integrated cost information tool. This system is able to extract data from the CAD-system, combine it with data about the company specific manufacturing environment, and subsequently autonomously evaluate manufacturability aspects and costs of the given CAD-model. Within minutes the designer gets quantitative in-formation about the major cost sources of his/her design. Additionally, some alternative methods for approximating manu-facturing times from empirical data, namely neural networks and local weighted regression, are introduced.

am

[BibTex]

[BibTex]