This chapter focuses on endowing robustness in the task execution step using vision‐based robot control. It presents a framework that performs all the steps of the learning process in the image space of a vision sensor. The developed approach for robot programming by demonstration (PbD) is based on the following postulations: perception of the demonstrations is performed with vision cameras; and execution of the learned strategies is conducted using visual feedback from the scene. Different camera configurations for observation of demonstrations have been used in the literature, for example, stereo pairs and multiple cameras. For generating the reference trajectories, instead of Kalman smoothing, other machine learning methods can be employed, for example, hidden Markov model (HMM), Gaussian mixture model (GMM)/Gaussian mixture regression (GMR), and dynamical systems approach. However, since the reference trajectories are subjected to correction for visual tracking, simple averaging via Kalman smoothing is adopted.
Task Execution
Robot Learning by Visual Observation ; 129-188
10.03.2017
60 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
SECOND TASK EXECUTION ASSISTANCE DEVICE AND SECOND TASK EXECUTION ASSISTANCE PROGRAM
Europäisches Patentamt | 2021
|TASK EXECUTION SUPPORT SYSTEM AND METHOD OF SUPPORTING TASKS EXECUTION
Europäisches Patentamt | 2019
|Europäisches Patentamt | 2024
|