Animal-Like Eye Vision Assisted Locomotion of a Quadruped Based on Reinforcement Learning.
Xiaojian WeiQing WeiHonglei AnZhitong ZhangJunwei YuHongxu MaPublished in: ICIRA (1) (2023)
Keyphrases
- reinforcement learning
- quadruped robot
- legged robots
- visual field
- real time
- robot control
- eye movements
- vision system
- function approximation
- image processing
- mobile robot
- learning algorithm
- computer vision
- multi agent
- real robot
- state space
- reinforcement learning algorithms
- eye tracking
- optimal policy
- model free
- rough terrain
- multi agent reinforcement learning
- multi agent systems
- learning classifier systems
- visual perception
- eye gaze
- markov decision processes
- dynamic programming
- visual attention
- legged locomotion
- smooth pursuit
- transition model
- gaze estimation
- learning process
- active vision
- learning problems
- temporal difference
- degrees of freedom