Acquiring robot skills via reinforcement learning

By Vijaykumar Gullapalli, Judy A. Franklin, Hamid Benbrahim. 1994

How researchers see this paper

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

Citations in the corpus

Learning concurrent motor skills in versatile solution spaces (2012, background): “Robot learning of dynamic motor tasks has been an active field of research in recent years and many impressive applications have been demonstrated [1], [2], [3], [4], [5].”
Finding Structure in Reinforcement Learning (1994, methodology): “In this paper, performance loss and description length has been defined based on table look-up representations of . Recently, various researchers have applied reinforcement learning in combination with generalizing function approximators, such as nearest neighbor methods or artificial neural network”
Reinforcement learning of motor skills with policy gradients (2008, methodology): “colleagues (Benbrahim & Franklin, 1997; Gullapalli, Franklin, & Benbrahim, 1994 ) in the early 1990s, these methods have been applied to a variety of robot learning problems ranging from simple control tasks (e.g., balancing a ball on a beam (Benbrahim, Doleac, Franklin, & Selfridge, 1992), and pol”
Evaluation of Policy Gradient Methods and Variants on the Cart-Pole Benchmark (2007, background): “The later requirement is particularly important as highly complex plants, e.g., a nthromorphic robot arms [ 4 ], [8] or legged robots [2], [6], [13], require highly task specific adaptation of both the pol icy and the algorithm.”
A lifelong learning perspective for mobile robot control (1994, background): “For example, in [5] a robot manipulator is described which learns to insert a peg in a hole without prior knowledge regarding the manipulator or the hole.”
A Review of Robot Learning for Manipulation: Challenges, Representations, and Algorithms (2019, background): “ntrol signals. However, the outputs of policies often do not work directly at the level of actuator signals. In practice, an additional controller is often placed between the policy and the actuator (Gullapalli et al., 1994). Using an additional controller allows the robot to leverage a large body o”
Policy Gradient Methods for Robotics (2006, background): “Starting with the pioneering work of Gullapali, Franklin and Benbrahim [1], [2] in the early 1990s, these methods have been applied to a variety of robot learning problems ranging from simple control tasks (e.g., balancing a ball-on a beam [3], and pole-balancing [4]) to complex learning tasks invol”