Learning and generalization of motor skills by learning from demonstration

By Peter Pastor, Heiko Hoffmann, Tamim Asfour, Stefan Schaal. 2009

827 citations recorded by the source corpus.

Subfields: Control and Systems Engineering, Mechanical Engineering, Reinforcement Learning

How researchers see this paper

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

In Control and Systems Engineering

In Mechanical Engineering

In Reinforcement Learning

Citations in the corpus

End-to-End Training of Deep Visuomotor Policies (2015, methodology): “The policy might specify a trajectory in task-space, relying on hand-designed PD controllers to execute the desired motion, and a policy for manipulating objects might rely on an existing vision system to localize these objects [29].”
Learning Parameterized Skills (2012, methodology): “Existing work has shown that it is possible to transfer information between pairs of related problems [76, 75, 44, 3] and that parameterized policies can be constructed to deal with slight variations of a domain [59, 53, 33, 27].”
Deep reinforcement learning for robotic manipulation with asynchronous off-policy updates (2016, background): “However, practical real-world applications of RL typically require significant engineering beyond the learning algorithm itself: an appropriate representation for the policy or value function must be chosen so as to achieve training times that are practical for physical hardware [8], and example dem”
Time-Contrastive Networks: Self-Supervised Learning from Video (2017, methodology): “robotics, this might be accomplished by means of kinesthetic demonstrations [11] or teleoperation [12], but both methods”
End-to-End Driving Via Conditional Imitation Learning (2017, methodology): “In robotics, movement primitives have been used as building blocks for advanced motor skills [17], [26].”
One-Shot Visual Imitation Learning via Meta-Learning (2017, background): “Efﬁcient imitation from a small number of demonstrations has been successful in scenarios where the state of the environment, such as the poses of objects, is known [8, 9, 10, 11].”
Deep spatial autoencoders for visuomotor learning (2015, methodology): “Reinforcement learning and policy search methods have been applied to perform a variety of robotic tasks, such as table tennis [9], object manipulation [10], [11], and locomotion [12], [13], [14], [15].”
An Algorithmic Perspective on Imitation Learning (2018, methodology): “planning literature, the term “via-point” is used similarly to the term keyframe [Pastor et al., 2009, Paraschos et al., 2013, Zucker et al., 2013].”
Learning Latent Plans from Play (2019, background): “Conventionally, obtaining multiple skills involves defining a discrete set of tasks we care about, collecting a large number of labeled and segmented expert demonstrations per task, then training one specialist policy per task in a learning from demonstration (LfD) [3] scenario.”
Path integral guided policy search (2016, methodology): “Policy search methods have been used in robotics for a variety of tasks, such as manipulation [4, 9, 10], playing table tennis [11] and ball-in-a-cup [3] games, regrasping [12], and locomotion [1, 2, 13].”
Learning contact-rich manipulation skills with guided policy search (2015, methodology): “Previous policy search methods in robotics generally use one or more of the following techniques to achieve good performance and reasonable training times: the use of example demonstrations [15], [9], [16], the use of a prior controller that allows the policy to only set high level targets, such as ”
One-Shot Imitation from Observing Humans via Domain-Adaptive Meta-Learning (2018, background): “acquire a range of complex skills through demonstration, such as table tennis [28], lane following [34], pouring water [31], drawer opening [38], and multi-stage manipulation tasks [62].”
Vision-Based Multi-Task Manipulation for Inexpensive Robots Using End-to-End Learning from Demonstration (2017, methodology): “In recent years, LfD was successfully used in the field of robotics for applications as varied as autonomous helicopter maneuvers [1], playing table tennis [2], object manipulation [3], making coffee [4], grasping novel objects [5], carrot grating [6], and eggplant cutting [7].”
Adapting Deep Visuomotor Representations with Weak Pairwise Constraints (2015, methodology): “rete state estimation system to distill the visual signal into a low-dimensional state (e.g., the positions of objects in the world). This state is then used as a noisy input into the control policy (Pastor et al., 2009). Recently, there has been considerable interest in learning visuomotor policies”
Learning Dexterous Manipulation Policies from Experience and Imitation (2016, methodology): “Indeed RL has been successfully applied to a range of robotic tasks (Tedrake, Zhang and Seung 2004; Kober, Oztop and Peters 2010; Pastor, Hoffmann, Asfour and Schaal 2009; Deisenroth, Rasmussen and Fox 2011), however dimensionality and sample complexity have presented major challenges (Kober,…”
A Divergence Minimization Perspective on Imitation Learning Methods (2019, methodology): “In many cases however, it is either too challenging to optimize a given reward (e.g. due to sparsity of signal), or it is simply impossible to design a reward function that captures the intricate details of desired outcomes [3, 4, 5, 6, 7, 8].”
BM: An iterative algorithm to learn stable non-linear dynamical systems with Gaussian mixture models (2010, background): “The recent modifications on DMP [14] do not address these issues.”
BC-Z: Zero-Shot Task Generalization with Robotic Imitation Learning (2022, neutral): “Imitation learning has been successful in learning grasping and pick-place tasks from low52 dimensional state [9, 10, 11, 12, 13, 14, 15].”