A Two-stage Learning Architecture that Generates High-Quality Grasps for a Multi-Fingered Hand

Keyphrases

• 6D Pose 50%
• Bayesian Optimization 50%
• Benchmark Dataset 50%
• Finger Joint 50%
• Finger Position 50%
• Grasp Planning 50%
• Grasp Pose 50%
• Grasp Quality 50%
• Grasping Unknown Objects 50%
• Hand Pose 50%
• High Dimensional Search Space 50%
• Joint Configuration 50%
• Learning Architecture 100%
• Learning Process 50%
• Multifingered Hands 100%
• Object Manipulation 50%
• Object Representation 100%
• Optimization Scheme 50%
• Physics-based 50%
• Planning Algorithms 50%
• Real Robots 50%
• Regress 50%
• Robot System 50%
• Robotic Hand 50%
• Scalar 50%
• Stable Grasp 100%
• Stage-wise 50%
• Success Rate 50%
• Synthetic Training Data 50%
• Two-way Learning 100%

Computer Science

• Learning Process 50%
• Object Representation 100%
• Planning Algorithm 50%
• robotic hand 50%
• Robotics 50%
• Search Space 50%
• Training Data 50%