MAT: Multi-Fingered Adaptive Tactile Grasping via Deep Reinforcement Learning.

Bohan Wu Iretiayo Akinola Jacob Varley Peter K. Allen

Published in: CoRL (2019)

Keyphrases

reinforcement learning
object manipulation
control theoretic
robot control
learning capabilities
case study
evolutionary algorithm
function approximation
neural network
state space
vision system
multi agent
dynamic programming
supervised learning
transfer learning
robotic control