H-InDex: Visual Reinforcement Learning with Hand-Informed Representations for Dexterous Manipulation.
Yanjie ZeYuyao LiuRuizhe ShiJiaxin QinZhecheng YuanJiashun WangHuazhe XuPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- haptic feedback
- manipulation tasks
- visual information
- human hand
- visual representations
- object manipulation
- visual features
- visual feedback
- markov decision processes
- model free
- visual perception
- visual patterns
- function approximation
- database
- mid level
- real robot
- optimal policy
- virtual environment
- supervised learning
- learning algorithm
- machine learning
- visual representation
- finite element analysis
- multiple representations
- low level