Bridging the Last Mile in Sim-to-Real Robot Perception via Bayesian Active Learning.
Jianxiang FengJongseok LeeMaximilian DurnerRudolph TriebelPublished in: CoRR (2021)
Keyphrases
- real robot
- active learning
- mobile robot
- simulated robot
- reinforcement learning
- learning strategies
- real world
- semi supervised
- active exploration
- supervised learning
- multi robot
- robotic systems
- bayesian networks
- random sampling
- machine learning
- humanoid robot
- learning algorithm
- robot soccer
- selective sampling
- labeled data
- relevance feedback
- motion control
- manipulation tasks
- path planning
- maximum likelihood
- training set
- experimental design
- real environment
- robot motion
- evolutionary robotics