PlayVirtual: Augmenting Cycle-Consistent Virtual Trajectories for Reinforcement Learning.
Tao YuCuiling LanWenjun ZengMingxiao FengZhibo ChenPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- function approximation
- virtual environment
- multi agent
- learning algorithm
- state space
- virtual world
- virtual reality
- augmented reality
- moving object trajectories
- machine learning
- reinforcement learning algorithms
- model free
- temporal difference
- markov decision processes
- learning capabilities
- trajectory data
- consistency constraints
- function approximators
- policy search
- optimal policy