Multi-task policy search for robotics

By Marc Peter Deisenroth, Péter Englert, Jan Peters, Dieter Fox. 2014

138 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Control and Systems Engineering, Reinforcement Learning, 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 Reinforcement Learning

In Robotics & Control

Citations in the corpus

Learning Parameterized Skills (2012, methodology): “Our method is agnostic to policy and skill representation and can be coupled with any of the recently-proposed parameterized skill learning algorithms [33, 17, 51, 19].”
RoboNet: Large-Scale Multi-Robot Learning (2019, background): “The multi-task literature [18, 19], lifelong learning literature [20, 21], and meta-learning literature [22, 23] describe ideas that are tightly coupled with this concept.”
End-to-End Driving Via Conditional Imitation Learning (2017, methodology): “Parameterized goals have been used to train motion controllers in robotics [7], [8], [18].”
One-Shot Visual Imitation Learning via Meta-Learning (2017, methodology): “Unlike prior methods that take the task identity [1, 2, 3, 4] or a demonstration [5] as the input into a contextual policy, our approach learns a parameterized policy that can be adapted to different tasks through gradient updates, effectively learning to imitation learn.”
An Algorithmic Perspective on Imitation Learning (2018, methodology): “Model-Based Behavioral Cloning Methods 105 [Englert et al., 2013, Deisenroth et al., 2014].”
Vision-Based Multi-Task Manipulation for Inexpensive Robots Using End-to-End Learning from Demonstration (2017, background): “It is possible to learn this kind of policies by considering both the state and the task as input to the policy [24].”
Meta Reinforcement Learning with Latent Variable Gaussian Processes (2018, background): “In [8], the authors consider the same dynamical system but solve different tasks by augmenting the policy with a task variable.”
Training Agents using Upside-Down Reinforcement Learning (2019, background): “There is also substantial prior work on learning reward or goal-conditional policies that directly produce actions [5, 6, 15, 35], but these rely either on a pre-trained world model, a dataset of goal and optimal policy parameters, or policy gradients for learning.”
Hindsight policy gradients (2017, background): “Recent examples of learning goal-conditional behavior include the works of Da Silva et al. [3], Schmidhuber [4], Srivastava et al. [5], Kupcsik et al. [6], Deisenroth et al. [7], Fabisch and Metzen [8], Schaul et al. [9], Zhu et al. [10], Held et al. [11].”
Learning Deep Parameterized Skills from Demonstration for Re-targetable Visuomotor Control (2019, methodology): “Parameterized skill learning as described by da Silva et al. [4] aims to learn a mapping from a given task parameter vector to a policy [4, 5, 6, 9, 16, 17, 21, 22, 24].”
MT-Opt: Continuous Multi-Task Robotic Reinforcement Learning at Scale (2021, methodology): “Combining multiple task policies has been explored in reinforcement learning by using gating networks [76, 50], conditioning policies on tasks [15], mapping tasks to parameters of a policy [13, 38, 82], distilling separate task policies into a shared multi-task policy [40, 77, 62, 53, 27, 5].”
Model-based contextual policy search for data-efficient generalization of robot skills (2017, methodology): “Recently, the PILCO algorithm has been extended for learning generalized lower-level controllers [11].”
Distributed Multitask Reinforcement Learning with Quadratic Convergence (2018, background): “Applications with these characteristics are ubiquitous ranging from stock-trading [1] to robotics control [2, 3].”
Automatic Goal Generation for Reinforcement Learning Agents (2017, background): “As opposed to the work of Deisenroth et al. (2014), our work uses a curriculum to perform efﬁcient multi-task learning, even in sparse reward settings.”
RT-1: Robotics Transformer for Real-World Control at Scale (2022, background): “set or some other parameterized form [7, 8, 13, 29].”