Deep learning can accelerate grasp-optimized motion planning.
Jeffrey IchnowskiYahav AvigalVishal SatishKen GoldbergPublished in: Sci. Robotics (2020)
Keyphrases
- motion planning
- deep learning
- degrees of freedom
- mobile robot
- path planning
- trajectory planning
- unsupervised feature learning
- unsupervised learning
- humanoid robot
- deep architectures
- machine learning
- weakly supervised
- multi robot
- robotic tasks
- mental models
- collision free
- climbing robot
- spatio temporal
- mean shift
- multi modal
- text mining
- data points