Abstract
Most objects are composed of semantically distinctive parts that are more or less geometrically distinctive as well. Points on the object relevant for a certain robot operation are usually determined by various physical properties of the object, such as its dimensions or weight distribution, and by the purpose of object parts. A robot operation defined for a particular part of a representative object can be transferred and adapted to other instances of the same object class by detecting the corresponding components. In this paper, a method for semantic association of the object’s components within the object class is proposed. It is suitable for real-time robotic tasks and requires only a few previously annotated representative models. The proposed approach is based on the component association graph and a novel descriptor that describes the geometrical arrangement of the components. The method is experimentally evaluated on a challenging benchmark dataset.
Highlights
One of the trends in robotics is to reduce the need for robot programing by allowing a robot to learn certain tasks from a human instructor
The aim of the paper was to investigate how accurate semantic association can be achieved without extensive training and annotation of a large amount of data and if such an approach could be time effective for real-time applications
The target application of the proposed approach was in robotics, where it could be used in combination with an object detection and 3D reconstruction module
Summary
One of the trends in robotics is to reduce the need for robot programing by allowing a robot to learn certain tasks from a human instructor. One approach to this problem is kinesthetic training of a robot, where a human manually guides a robot manipulator to perform certain action and the robot applies the learned action to solve a practical task [1,2]. An advanced version of such training would be to define a robot action for a particular instance of an object class, referred to in this paper as a representative object, and apply an algorithm that would adapt this action to the other instances of the same class. Since the target application field considered in this paper is robotics, we define components as regions of the object’s surface that could potentially represent contact surfaces between the object and a robot tool when performing some task
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
