Online Goal Babbling for rapid bootstrapping of inverse models in high dimensions

Reimann, Matthias; Steil, Jochen J.; Gienger, Michael

doi:10.1109/devlrn.2011.6037368

Cited by 69 publications

(101 citation statements)

References 22 publications

Supporting

Mentioning

101

Contrasting

Order By: Relevance

“…When viewed in the SSP setting, each topic can be either a local predictive forward model [20], [43], an option [22], or a region in a parameterized goal/option space [11], i.e. a local inverse model (see also [44] and [45] for similar ideas). In these works the tasks cannot be sampled simultaneously and a decision has to be made.…”

Section: A Multi-topicmentioning

confidence: 99%

The strategic student approach for life-long exploration and learning

Lopes

Oudeyer

2012

2012 IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL)

View full text Add to dashboard Cite

Abstract-This article introduces the strategic student metaphor: a student has to learn a number of topics (or tasks) to maximize its mean score, and has to choose strategically how to allocate its time among the topics and/or which learning method to use for a given topic. We show that under which conditions a strategy where time allocation or learning method is chosen from the easier to the more complex topic is optimal. Then, we show an algorithm, based on multi-armed bandit techniques, that allows empirical online evaluation of learning progress and approximates the optimal solution under more general conditions. Finally, we show that the strategic student problem formulation allows to view in a common framework many previous approaches to active and developmental learning.

show abstract

Section: A Multi-topicmentioning

confidence: 99%

The strategic student approach for life-long exploration and learning

Lopes

Oudeyer

2012

2012 IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL)

View full text Add to dashboard Cite

show abstract

“…Additionally, it was shown in previous models that learning redundant inverse models could be achieved more efficiently if exploration was driven by goal babbling, triggering reaching, rather than driven by direct motor babbling [5], [16].…”

Section: Introductionmentioning

confidence: 99%

Curiosity-driven phonetic learning

Moulin-Frier

Oudeyer

2012

2012 IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL)

View full text Add to dashboard Cite

Abstract-This article studies how developmental phonetic learning can be guided by pure curiosity-driven exploration, also called intrinsically motivated exploration. Phonetic learning refers here to learning how to control a vocal tract to reach acoustic goals. We compare three different exploration strategies for learning the auditory-motor inverse model: random motor exploration, random goal selection with reaching, and curiositydriven active goal selection with reaching. Using a realistic vocal tract model, we show how intrinsically motivated learning driven by competence progress can generate automatically developmental structure in both articulatory and auditory modalities, displaying patterns in line with some experimental data from infants.

show abstract

“…Goal babbling is based on work by Rolf et al [19], where the comparison is made to infants who tend to make goaldirected movements from at a very early age even if they do not succeed. Accordingly only a smaller relevant subspace of possible motor commands is explored, as opposed to random motor babbling.…”

Section: A Goal Babblingmentioning

confidence: 99%

“…In [19], the inverse kinematics is learned in 2D space by online construction of local linear regressions over positions that are weighted by prototype vectors. A similar local weighting is discussed in more detail in Sect.…”

Section: A Goal Babblingmentioning

confidence: 99%

“…In order to deal with the increasing difficulty of learning the inverse dynamics of musculoskeletal robots driven by PAMs, in this work, we combine a variant of goal babbling [19] with an echo state network (ESN) [9], a particular kind of recurrent neural network, to represent the state of the system. Gaussian process regression (GPR) is used to train the network in realtime (see Sects.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Real-Time Inverse Dynamics Learning for Musculoskeletal Robots based on Echo State Gaussian Process Regression

Hartmann¹,

Boedecker²,

Obst³

et al. 2012

Robotics: Science and Systems VIII

View full text Add to dashboard Cite

Abstract-A challenging topic in articulated robots is the control of redundantly many degrees of freedom with artificial muscles. Actuation with these devices is difficult to solve because of nonlinearities, delays and unknown parameters such as friction. Machine learning methods can be used to learn control of these systems, but are faced with the additional problem that the size of the search space prohibits full exploration in reasonable time. We propose a novel method that is able to learn control of redundant robot arms with artificial muscles online from scratch using only the position of the end effector, without using any joint positions, accelerations or an analytical model of the system or the environment. To learn in real time, we use the so called online "goal babbling" method to effectively reduce the search space, a recurrent neural network to represent the state of the robot arm, and novel online Gaussian processes for regression. With our approach, we achieve good performance on trajectory tracking tasks for the end effector of two very challenging systems: a simulated 6 DOF redundant arm with artificial muscles, and a 7 DOF robot arm with McKibben pneumatic artificial muscles. We also show that the combination of techniques we propose results in significantly improved performance over using the individual techniques alone.

show abstract

Online Goal Babbling for rapid bootstrapping of inverse models in high dimensions

Cited by 69 publications

References 22 publications

The strategic student approach for life-long exploration and learning

The strategic student approach for life-long exploration and learning

Curiosity-driven phonetic learning

Real-Time Inverse Dynamics Learning for Musculoskeletal Robots based on Echo State Gaussian Process Regression

Contact Info

Product

Resources

About