Learning perceptual coupling for motor primitives

By Jens Kober, Betty J. Mohler, Jan Peters. 2008

92 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Cognitive Neuroscience, Control and Systems Engineering

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Cognitive Neuroscience

In Control and Systems Engineering

Citations in the corpus

Reinforcement learning in robotics: A survey (2013, methodology): “Non-parametric Kober et al. (2010); Mitsunaga et al.”
An Algorithmic Perspective on Imitation Learning (2018, methodology): “Kohlmorgen and Lemm [2001] developed an online segmentation method based on HMMs. By computing the “distance" between nearby data windows, Kohlmorgen and Lemm [2001] segments human motion data using unsupervised learning.”
Path integral guided policy search (2016, methodology): “Reinforcement learning (RL) and policy search methods have shown considerable promise for enabling robots to automatically learn a wide range of complex skills [1, 2, 3, 4], and recent results in deep reinforcement learning suggest that this capability can be extended to learn nonlinear policies tha”
Active Reward Learning (2014, background): “It does not rely on exhaustive sampling of the state-action space and, as a result, many recent advances in robot learning rely on policy search [22, 20, 26].”
Active reward learning with a novel acquisition function (2015, background): “It does not rely on exhaustive sampling of the state-action space and, as a result, many recent advances in robot learning are based on PS (Kormushev et al. 2010; Kober et al. 2008; Ng and Coates 1998).”
Movement templates for learning of hitting and batting (2010, background): “Additional feedback terms can be added as shown in [3], [16], [17].”
Generalization of human grasping for multi-fingered robot hands (2012, background): “DMPs have been widely adopted in the robotics community, and are well-known for their use in imitation learning [21], [18].”
Deriving and improving CMA-ES with information geometric trust regions (2017, methodology): “Subsequently, a new search distribution is computed by either computing gradient based updates [19], expectation-maximisationbased updates [7, 12], evolutionary strategies [10], the cross-entropy method [14] or information-theoretic policy updates [1], such that the individuals with higher fitness w”
Modeling Long-horizon Tasks as Sequential Interaction Landscapes (2020, background): “grasping) for completing tasks and a number of approaches exist that focus on the parallel activation and the smooth blending of movement primitives [10, 22, 23, 24, 25].”
Towards Motor Skill Learning for Robotics (2007, methodology): “From this point of view, one can straightforwardly derive regression algorithms such as the ones in [4, 14] to imitate using both the standard formulation of motor primitives [4] as well as the perceptually coupled formulation [14].”
Imitation and Reinforcement Learning (2010, background): “A series of such frameworks has been introduced to date [2], [3].”
Hierarchical Relative Entropy Policy Search (2014, methodology): “We are using the extended version of DMPs by Kober et al. (2010a), that parametrizes trajectories using a combination of weighted basis functions as well as the end point”
Hierarchical visuomotor control of humanoids (2018, background): “Particularly relevant to our cause is the work of Kober et al. (2008) in which motor primitives were coupled by reinforcement learning to external perceptual inputs.”
Imitation Learning (2017, background): “…with a dynamic system of motor primitives has produced a number of applications for learning discrete and rhythmic motions [Ijspeert et al. 2002a; Schaal et al. 2007; Kober et al. 2008], though most approaches focus on direct imitation without further optimization [Kober and Peters 2009a].”
Learning to select and generalize striking movements in robot table tennis (2012, methodology): “Here, learning from demonstrations can provide a good starting point for motor skill learning as it allows the efficient acquisition of single movement primitives (Guenter et al., 2007; Peters and Schaal, 2008b; Kober et al., 2008; Peters and Schaal, 2008a; Bitzer et al., 2010).”
Reinforcement learning for robot soccer (2009, background): “2004), teaching a robot to hit a baseball (Peters and Schaal 2006), operational space control (Peters and Schaal 2008a) and learning ball-in-a-cup with a robot arm (Kober et al. 2008).”
Stable reinforcement learning with autoencoders for tactile and visual data (2016, background): “A popular policy parametrization uses dynamic motor primitives to generate trajectories to be tracked [4], [5], [11]–[14].”
Learning motor primitives for robotics (2009, methodology): “A series of such frameworks has been introduced to date [2]–[5].”
A Survey on Policy Search for Robotics (2013, methodology): “This experiment was extended in [36] to include feedback for the position of the ball.”