DeepWalk: Omnidirectional Bipedal Gait by Deep Reinforcement Learning.
Diego RodriguezSven BehnkePublished in: ICRA (2021)
Keyphrases
- reinforcement learning
- walking speed
- gait recognition
- mobile robot
- human gait
- temporal difference
- gait analysis
- function approximation
- human recognition
- stereo vision
- closed loop
- optimal control
- human identification
- reinforcement learning algorithms
- model free
- human gait recognition
- learning algorithm
- vision system
- markov decision processes
- humanoid robot
- single camera
- limit cycle
- autonomous robots
- optimal policy
- learning process
- real robot
- deep learning
- function approximators
- neural network
- person identification
- stereo matching
- transfer learning
- supervised learning
- dynamic programming
- egomotion estimation