From object categories to grasp transfer using probabilistic reasoning

In this paper we address the problem of grasp generation and grasp transfer between objects using categorical knowledge. The system is built upon an i) active scene segmentation module, able of generating object hypotheses and segmenting them from the background in real time, ii) object categorization system using integration of 2D and 3D cues, and iii) probabilistic grasp reasoning system. Individual object hypotheses are first generated, categorized and then used as the input to a grasp generation and transfer system that encodes task, object and action properties. The experimental evaluation compares individual 2D and 3D categorization approaches with the integrated system, and it demonstrates the usefulness of the categorization in task-based grasping and grasp transfer.

Authors

• Marianna Madry
• Dan Song 0002
• Danica Kragic

Keywords

• Cognition
• Robots
• Kernel
• Probable Reason
• Experimental Evaluation
• Object Properties
• Activation Segment
• Scene Segmentation
• Linear Method
• Point Cloud
• Nonlinear Method
• Object Shape
• 3D Point
• Natural Scenes
• Nonlinear Techniques
• Objects In The Scene
• Point Cloud Data
• Vision Sensors
• 2D Data
• Task Constraints
• Sum Rules
• Product Rule
• 2D Descriptors
• Contour Shape
• Single Cue
• Algebraic Method
• 3D Descriptors
• Point Cloud Segmentation
• Object Recognition
• 3D Shape
• System Performance
• Viewpoint Changes