Task-Specific Generalization of Discrete and Periodic Dynamic Movement Primitives

By Ales Ude, Andrej Gams, Tamim Asfour, Jun Morimoto. 2010

379 citations recorded by the source corpus.

Subfields: Biomedical Engineering, Computer Vision and Pattern Recognition, 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 Biomedical Engineering

In Computer Vision and Pattern Recognition

In Control and Systems Engineering

Citations in the corpus

A tutorial on task-parameterized movement learning and retrieval (2015, methodology): “Approaches employing M models for the M demonstrations, performed in M diﬀerent situations, see e.g. [29, 47, 59, 50, 97, 21, 44]; 2.”
Probabilistic Movement Primitives (2013, methodology): “MPs have been used to solve many complex tasks, including ‘Ball-in-the-Cup’ [4], Ball-Throwing [5, 6], Pancake-Flipping [7] and Tetherball [8].”
Learning position and orientation dynamics from demonstrations via contraction analysis (2018, methodology): “Tasks, such as drumming (Ude et al. 2010), pouring (Nemec et al. 2009), and pick and place have been recreated using DMPs on various robotic platforms.”
Learning control Lyapunov function to ensure stability of dynamical system-based robot reaching motions (2014, methodology): “DMP has been used for different robotics applications such as: walking [32], drumming [30], pouring [49], and obstacle avoidance [ 36, 37].”
Learning compact parameterized skills with a single regression (2013, methodology): “Training a DMP from multiple demonstrations is done with one single regression, rather than current approaches which require two separate regressions [5]–[8].”
Model-based contextual policy search for data-efficient generalization of robot skills (2017, background): “In [15, 40], robot skills are represented by Dynamic Movement Primitives [17] or DMPs.”
A probabilistic approach to robot trajectory generation (2013, methodology): “Typically, the generalization to new final positions or velocities is based on heuristics and cannot be directly obtained by imitation learning.”
Learning Partially Contracting Dynamical Systems from Demonstrations (2017, methodology): “Tasks such as drumming [29], pouring [30], and pick and place have been recreated using DMPs on various robotic platforms.”
Learning to select and generalize striking movements in robot table tennis (2012, methodology): “An exception is the work of Ude et al. (2010), in which the internal parameters of the DMPs are recomputed from a library of movement primitives in each trial using locally weighted regression.”
Learning robot motions with stable dynamical systems under diffeomorphic transformations (2015, background): “They are often used to provide a library of basic components called movement primitives (MP) [4, 5], which are applied very successfully to generate movements in a variety of manipulation tasks [6].”
A Dynamical System-based Approach to Modeling Stable Robot Control Policies via Imitation Learning (2012, methodology): “During the last decade, the IL-based approach to estimate DS model of robot motions has become increasingly popular as it offers a mechanism to exploit the advantages of both robot learning and DS modeling (Dixon & Khosla, 2004a; Kober, Mulling, et al., 2010; Ude et al., 2010; Hersch et al., 2008; K”
Neural learning of stable dynamical systems based on data-driven Lyapunov candidates (2013, background): “For instance, point-to-point movements modeled by autonomous dynamical systems can provide a library of basic building blocks called movement primitives [4] which are very successfully applied to generate movements in a variety of manipulation tasks [5], [6].”
Towards Robust Skill Generalization: Unifying Learning from Demonstration and Motion Planning (2017, background): “Attempts at combining probabilistic approaches with dynamical systems [9, 10, 11] have also met some success at generalization.”
Hierarchical Relative Entropy Policy Search (2014, methodology): “We treat the hierarchical structure of the policy as a latent variable, which results in the Hierarchical Relative Entropy Policy Search (HiREPS) algorithm.”
SWIRL: A sequential windowed inverse reinforcement learning algorithm for robot tasks with delayed rewards (2018, background): “2011), adaptation to unseen scenarios (Ijspeert et al. 2002; Ude et al. 2010), and demonstrator skill-assessment (Reiley et al.”
Survey on human–robot collaboration in industrial settings: Safety, intuitive interfaces and applications (2018, background): “However, most of works related to PbD consist in theoretical and experimental approaches, and appear far to be ready for everyday implementation in industrial practice [140, 141, 142, 143, 144].”