A Graph-Based Deep Reinforcement Learning Approach to Grasping Fully Occluded Objects.
Guoyu ZuoJiayuan TongZihao WangDaoxiong GongPublished in: Cogn. Comput. (2023)
Keyphrases
- occluded objects
- reinforcement learning
- d objects
- model free
- function approximation
- optimal policy
- deep learning
- multiple objects
- learning algorithm
- reinforcement learning algorithms
- markov decision processes
- graph theoretic
- neural network
- multi agent
- vision system
- semi supervised
- pairwise
- multi agent reinforcement learning
- object manipulation
- machine learning
- optimal control
- state space
- graph model
- mobile robot
- object recognition
- robot control