Dextrous Tactile In-Hand Manipulation Using a Modular Reinforcement Learning Architecture.
Johannes PitzLennart RöstelLeon SieversBerthold BäumlPublished in: ICRA (2023)
Keyphrases
- reinforcement learning
- management system
- highly flexible
- real time
- learning capabilities
- function approximation
- autonomous learning
- markov decision processes
- evolutionary learning
- master slave
- architectural design
- layered architecture
- machine learning
- temporal difference
- multi agent reinforcement learning
- haptic feedback
- modular neural networks
- action selection
- evaluation function
- software architecture
- multi agent
- data sets