Learning motor primitives for robotics

By Jens Kober, Jan Peters. 2009

268 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Control and Systems Engineering, Social Psychology, Robotics & Control

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 Control and Systems Engineering

In Social Psychology

In Robotics & Control

Citations in the corpus

End-to-End Training of Deep Visuomotor Policies (2015, methodology): “ples in each batch. With Gaussian trajectories, CEM and RWR only dier in the weights. These methods represent a class of RL algorithms that t the policy to weighted samples, including PoWER and PI2 (Kober and Peters, 2009; Theodorou et al., 2010; Stulp and Sigaud, 2012). PILCO is a model-based metho”
Deep reinforcement learning for robotic manipulation with asynchronous off-policy updates (2016, background): “Of course, many other linear representations are possible, including DMPs [39], splines [3], and task-specific representations [40].”
Reinforcement learning in robotics: A survey (2013, methodology): “…Atkeson and Schaal (1997), Bentivegna et al. (2004), Bitzer et al. (2010), Conn and Peters II (2007), Gräve et al. (2010), Kober et al. (2008); Kober and Peters (2009), Latzke et al. (2007), and Peters and Schaal (2008c,b) Policy Birdwell and Livingston (2007), Erden and Leblebicioğlu…”
Learning to Generalize Across Long-Horizon Tasks from Human Demonstrations (2020, background): “Imitation learning has been applied to multiple domains such as playing table tennis [22], Go [37] and video games [33], and driving autonomously [4, 31] In robotics, the idea of robots learning from human demonstrations have been extensively explored [36, 17, 23, 11, 14, 3, 5].”
Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations (2017, methodology): “d) Combining RL with demonstrations: Methods based on dynamic movement primitives (DMPs) [36], [50], [21], [18] have been used to effectively combine demonstrations and RL to enable faster learning.”
An Algorithmic Perspective on Imitation Learning (2018, background): “Moreover, application examples—such as robots that play table tennis [Kober and Peters, 2009] and programs that play the game of Go [Silver et al., 2016]— illustrate the properties and motivations behind different forms of imitation learning.”
Robot motor skill coordination with EM-based Reinforcement Learning (2010, methodology): “In order to minimize the number of rollouts which are needed to estimate new policy parameters, we use a form of importance sampling technique adapted for RL [3], [10] and”
Preventing undesirable behavior of intelligent machines (2019, methodology): “S15, here we propose a new quasi-Seldonian algorithm that takes a non-standard, but not uncommon [138, 139, 140, 141], approach to RL: we propose an algorithm that searches the space of probability distributions over policies rather than an algorithm like SPI that directly searches the space of poli”
Learning Complex Neural Network Policies with Trajectory Optimization (2014, methodology): “We also compared to cost-weighted regression, which fits the policy to previous on-policy samples weighted by the exponential of their reward (negative cost) (Peters & Schaal, 2007; Kober & Peters, 2009).”
Learning Robot Skills with Temporal Variational Inference (2020, methodology): “Sequencing Primitives: The concept of learning movement primitives using predefined representations of primitives (Kober & Peters, 2009; Peters et al., 2013) has been a popular technique to capturing robotic behaviors.”
Variational Policy Search via Trajectory Optimization (2013, unknown): “This is the standard formulation for expectation maximization [9], and has been applied to policy optimization in previous work [8, 21, 3, 11].”
Imitation from Observation: Learning to Imitate Behaviors from Raw Video via Context Translation (2017, background): “Imitation learning has enabled the successful performance of tasks in a number of complex domains such as helicopter flight through apprenticeship learning [12], learning how to put a ball in a cup and playing table tennis [13], performing human-like reaching”
Learning deep neural network policies with continuous memory states (2015, methodology): “Besides guided policy search, we evaluated the memory states approach with the reward-weighted regression (RWR) algorithm [23], [24].”
SkyAI: Highly modularized reinforcement learning library (2010, methodology): “config ={ Factor ={ resize(17, 5) [0]= ( 0, 0, 0, 0, 0) [1]= ( 1, 0, 0, 0, 0) [2]= ( 0, 0, 0, 0, 0) [3]= ( 0, 1, 0, 0, 0) [4]= ( 1, 0, 0, 0, 0) [5]= ( 0, 0, 0, 0, 0) [6]= ( 0, 1, 0, 0, 0) [7]= ( 0, 0, 0, 0, 0) [8]= ( 0, 0, 1, 0, 0) [9]= ( 0, 0, 0, 1, 0) [10]= ( 0, 0, 0, 0, 1) [11]= ( 0, 0, 0, 0, 0) ”
Discovering Motor Programs by Recomposing Demonstrations (2020, methodology): “…of techniques exist to do so, including cloning the demonstrated behavior (Esmaili et al., 1995), ﬁtting a parametric model to the demonstrations (Kober & Peters, 2009; Peters et al., 2013), or ﬁrst segmenting the demonstrations and ﬁtting a model to each of the resultant segments (Niekum et…”
Catch & Carry (2019, methodology): “A fairly direct approach involves fitting the demonstrations to a parametric form and using RL to modulate the parameters of the fitted model [Guenter et al. 2007; Peters and Schaal 2008; Kober and Peters 2009; Pastor et al. 2011].”
Learning Neural Network Policies with Guided Policy Search under Unknown Dynamics (2014, methodology): “With Gaussian trajectories, CEM and RWR only differ in the weights.”
A Machine Learning Approach to Visual Perception of Forest Trails for Mobile Robots (2016, methodology): “Ross et al. [3] used imitation learning [26], [27] to steer a quadrotor to avoid trees in an outdoor environment; the controller is previously trained by manually piloting the robot for a short time.”